Summary:
There's no technological impediment to allowing the Bloom
filter bits/key to be non-integer (fractional/decimal) values, and it
provides finer control over the memory vs. accuracy trade-off. This is
especially handy in using the format_version=5 Bloom filter in place
of the old one, because bits_per_key=9.55 provides the same accuracy as
the old bits_per_key=10.
This change not only requires refining the logic for choosing the best
num_probes for a given bits/key setting, it revealed a flaw in that logic.
As bits/key gets higher, the best num_probes for a cache-local Bloom
filter is closer to bpk / 2 than to bpk * 0.69, the best choice for a
standard Bloom filter. For example, at 16 bits per key, the best
num_probes is 9 (FP rate = 0.0843%) not 11 (FP rate = 0.0884%).
This change fixes and refines that logic (for the format_version=5
Bloom filter only, just in case) based on empirical tests to find
accuracy inflection points between each num_probes.
Although bits_per_key is now specified as a double, the new Bloom
filter converts/rounds this to "millibits / key" for predictable/precise
internal computations. Just in case of unforeseen compatibility
issues, we round to the nearest whole number bits / key for the
legacy Bloom filter, so as not to unlock new behaviors for it.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6092
Test Plan: unit tests included
Differential Revision: D18711313
Pulled By: pdillinger
fbshipit-source-id: 1aa73295f152a995328cb846ef9157ae8a05522a
Summary:
As described in detail in issue https://github.com/facebook/rocksdb/issues/6048, iterators' dereference operators
(`*`, `->`, and `[]`) should return `pointer`s/`reference`s (as opposed to
`const_pointer`s/`const_reference`s) even if the iterator itself is `const`
to be in sync with the standard's iterator concept.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6057
Test Plan: make check
Differential Revision: D18623235
Pulled By: ltamasi
fbshipit-source-id: 04e82d73bc0c67fb0ded018383af8dfc332050cc
Summary:
This is a required operator for random-access iterators, and an upcoming update for Visual Studio 2019 will change the C++ Standard Library's heap algorithms to use this operator.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6047
Differential Revision: D18618531
Pulled By: ltamasi
fbshipit-source-id: 08d10bc85bf2dbc3f7ef0fa3c777e99f1e927ef5
Summary:
Adds an improved, replacement Bloom filter implementation (FastLocalBloom) for full and partitioned filters in the block-based table. This replacement is faster and more accurate, especially for high bits per key or millions of keys in a single filter.
Speed
The improved speed, at least on recent x86_64, comes from
* Using fastrange instead of modulo (%)
* Using our new hash function (XXH3 preview, added in a previous commit), which is much faster for large keys and only *slightly* slower on keys around 12 bytes if hashing the same size many thousands of times in a row.
* Optimizing the Bloom filter queries with AVX2 SIMD operations. (Added AVX2 to the USE_SSE=1 build.) Careful design was required to support (a) SIMD-optimized queries, (b) compatible non-SIMD code that's simple and efficient, (c) flexible choice of number of probes, and (d) essentially maximized accuracy for a cache-local Bloom filter. Probes are made eight at a time, so any number of probes up to 8 is the same speed, then up to 16, etc.
* Prefetching cache lines when building the filter. Although this optimization could be applied to the old structure as well, it seems to balance out the small added cost of accumulating 64 bit hashes for adding to the filter rather than 32 bit hashes.
Here's nominal speed data from filter_bench (200MB in filters, about 10k keys each, 10 bits filter data / key, 6 probes, avg key size 24 bytes, includes hashing time) on Skylake DE (relatively low clock speed):
$ ./filter_bench -quick -impl=2 -net_includes_hashing # New Bloom filter
Build avg ns/key: 47.7135
Mixed inside/outside queries...
Single filter net ns/op: 26.2825
Random filter net ns/op: 150.459
Average FP rate %: 0.954651
$ ./filter_bench -quick -impl=0 -net_includes_hashing # Old Bloom filter
Build avg ns/key: 47.2245
Mixed inside/outside queries...
Single filter net ns/op: 63.2978
Random filter net ns/op: 188.038
Average FP rate %: 1.13823
Similar build time but dramatically faster query times on hot data (63 ns to 26 ns), and somewhat faster on stale data (188 ns to 150 ns). Performance differences on batched and skewed query loads are between these extremes as expected.
The only other interesting thing about speed is "inside" (query key was added to filter) vs. "outside" (query key was not added to filter) query times. The non-SIMD implementations are substantially slower when most queries are "outside" vs. "inside". This goes against what one might expect or would have observed years ago, as "outside" queries only need about two probes on average, due to short-circuiting, while "inside" always have num_probes (say 6). The problem is probably the nastily unpredictable branch. The SIMD implementation has few branches (very predictable) and has pretty consistent running time regardless of query outcome.
Accuracy
The generally improved accuracy (re: Issue https://github.com/facebook/rocksdb/issues/5857) comes from a better design for probing indices
within a cache line (re: Issue https://github.com/facebook/rocksdb/issues/4120) and improved accuracy for millions of keys in a single filter from using a 64-bit hash function (XXH3p). Design details in code comments.
Accuracy data (generalizes, except old impl gets worse with millions of keys):
Memory bits per key: FP rate percent old impl -> FP rate percent new impl
6: 5.70953 -> 5.69888
8: 2.45766 -> 2.29709
10: 1.13977 -> 0.959254
12: 0.662498 -> 0.411593
16: 0.353023 -> 0.0873754
24: 0.261552 -> 0.0060971
50: 0.225453 -> ~0.00003 (less than 1 in a million queries are FP)
Fixes https://github.com/facebook/rocksdb/issues/5857
Fixes https://github.com/facebook/rocksdb/issues/4120
Unlike the old implementation, this implementation has a fixed cache line size (64 bytes). At 10 bits per key, the accuracy of this new implementation is very close to the old implementation with 128-byte cache line size. If there's sufficient demand, this implementation could be generalized.
Compatibility
Although old releases would see the new structure as corrupt filter data and read the table as if there's no filter, we've decided only to enable the new Bloom filter with new format_version=5. This provides a smooth path for automatic adoption over time, with an option for early opt-in.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/6007
Test Plan: filter_bench has been used thoroughly to validate speed, accuracy, and correctness. Unit tests have been carefully updated to exercise new and old implementations, as well as the logic to select an implementation based on context (format_version).
Differential Revision: D18294749
Pulled By: pdillinger
fbshipit-source-id: d44c9db3696e4d0a17caaec47075b7755c262c5f
Summary:
For upcoming new SST filter implementations, we will use a new
64-bit hash function (XXH3 preview, slightly modified). This change
updates hash.{h,cc} for that change, adds unit tests, and out-of-lines
the implementations to keep hash.h as clean/small as possible.
In developing the unit tests, I discovered that the XXH3 preview always
returns zero for the empty string. Zero is problematic for some
algorithms (including an upcoming SST filter implementation) if it
occurs more often than at the "natural" rate, so it should not be
returned from trivial values using trivial seeds. I modified our fork
of XXH3 to return a modest hash of the seed for the empty string.
With hash function details out-of-lines in hash.h, it makes sense to
enable XXH_INLINE_ALL, so that direct calls to XXH64/XXH32/XXH3p
are inlined. To fix array-bounds warnings on some inline calls, I
injected some casts to uintptr_t in xxhash.cc. (Issue reported to Yann.)
Revised: Reverted using XXH_INLINE_ALL for now. Some Facebook
checks are unhappy about #include on xxhash.cc file. I would
fix that by rename to xxhash_cc.h, but to best preserve history I want
to do that in a separate commit (PR) from the uintptr casts.
Also updated filter_bench for this change, improving the performance
predictability of dry run hashing and adding support for 64-bit hash
(for upcoming new SST filter implementations, minor dead code in the
tool for now).
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5984
Differential Revision: D18246567
Pulled By: pdillinger
fbshipit-source-id: 6162fbf6381d63c8cc611dd7ec70e1ddc883fbb8
Summary:
This change sets up for alternate implementations underlying
BloomFilterPolicy:
* Refactor BloomFilterPolicy and expose in internal .h file so that it's easy to iterate over / select implementations for testing, regardless of what the best public interface will look like. Most notably updated db_bloom_filter_test to use this.
* Hide FullFilterBitsBuilder from unit tests (alternate derived classes planned); expose the part important for testing (CalculateSpace), as abstract class BuiltinFilterBitsBuilder. (Also cleaned up internally exposed interface to CalculateSpace.)
* Rename BloomTest -> BlockBasedBloomTest for clarity (despite ongoing confusion between block-based table and block-based filter)
* Assert that block-based filter construction interface is only used on BloomFilterPolicy appropriately constructed. (A couple of tests updated to add ", true".)
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5967
Test Plan: make check
Differential Revision: D18138704
Pulled By: pdillinger
fbshipit-source-id: 55ef9273423b0696309e251f50b8c1b5e9ec7597
Summary:
filter_bench is a specialized micro-benchmarking tool that
should not be needed with ROCKSDB_LITE. This should fix the LITE build.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5978
Test Plan: make LITE=1 check
Differential Revision: D18177941
Pulled By: pdillinger
fbshipit-source-id: b73a171404661e09e018bc99afcf8d4bf1e2949c
Summary:
* Adds support for plain table filter. This is not critical right now, but does add a -impl flag that will be useful for new filter implementations initially targeted at block-based table (and maybe later ported to plain table)
* Better mixing of inside vs. outside queries, for more realism
* A -best_case option handy for implementation tuning inner loop
* Option for whether to include hashing time in dry run / net timings
No modifications to production code, just filter_bench.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5968
Differential Revision: D18139872
Pulled By: pdillinger
fbshipit-source-id: 5b09eba963111b48f9e0525a706e9921070990e8
Summary:
Some filtering tests were unfriendly to new implementations of
FilterBitsBuilder because of dynamic_cast to FullFilterBitsBuilder. Most
of those have now been cleaned up, worked around, or at least changed
from crash on dynamic_cast failure to individual test failure.
Also put some clarifying comments on filter-related APIs.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5960
Test Plan: make check
Differential Revision: D18121223
Pulled By: pdillinger
fbshipit-source-id: e83827d9d5d96315d96f8e25a99cd70f497d802c
Summary:
- Updated our included xxhash implementation to version 0.7.2 (== the latest dev version as of 2019-10-09).
- Using XXH_NAMESPACE (like other fb projects) to avoid potential name collisions.
- Added fastrange64, and unit tests for it and fastrange32. These are faster alternatives to hash % range.
- Use preview version of XXH3 instead of MurmurHash64A for NPHash64
-- Had to update cache_test to increase probability of passing for any given hash function.
- Use fastrange64 instead of % with uses of NPHash64
-- Had to fix WritePreparedTransactionTest.CommitOfDelayedPrepared to avoid deadlock apparently caused by new hash collision.
- Set default seed for NPHash64 because specifying a seed rarely makes sense for it.
- Removed unnecessary include xxhash.h in a popular .h file
- Rename preview version of XXH3 to XXH3p for clarity and to ease backward compatibility in case final version of XXH3 is integrated.
Relying on existing unit tests for NPHash64-related changes. Each new implementation of fastrange64 passed unit tests when manipulating my local build to select it. I haven't done any integration performance tests, but I consider the improved performance of the pieces being swapped in to be well established.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5909
Differential Revision: D18125196
Pulled By: pdillinger
fbshipit-source-id: f6bf83d49d20cbb2549926adf454fd035f0ecc0d
Summary:
The parts that are used to implement FilterPolicy /
NewBloomFilterPolicy and not used other than for the block-based table
should be consolidated under table/block_based/filter_policy*.
This change is step 2 of 2:
mv util/bloom.cc table/block_based/filter_policy.cc
This gets its own PR so that git has the best chance of following the
rename for blame purposes. Note that low-level shared implementation
details of Bloom filters remain in util/bloom_impl.h, and
util/bloom_test.cc remains where it is for now.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5966
Test Plan: make check
Differential Revision: D18124930
Pulled By: pdillinger
fbshipit-source-id: 823bc09025b3395f092ef46a46aa5ba92a914d84
Summary:
The parts that are used to implement FilterPolicy /
NewBloomFilterPolicy and not used other than for the block-based table
should be consolidated under table/block_based/filter_policy*. I don't
foresee sharing these APIs with e.g. the Plain Table because they don't
expose hashes for reuse in indexing.
This change is step 1 of 2:
(a) mv table/full_filter_bits_builder.h to
table/block_based/filter_policy_internal.h which I expect to expand
soon to internally reveal more implementation details for testing.
(b) consolidate eventual contents of table/block_based/filter_policy.cc
in util/bloom.cc, which has the most elaborate revision history
(see step 2 ...)
Step 2 soon to follow:
mv util/bloom.cc table/block_based/filter_policy.cc
This gets its own PR so that git has the best chance of following the
rename for blame purposes. Note that low-level shared implementation
details of Bloom filters are in util/bloom_impl.h.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5963
Test Plan: make check
Differential Revision: D18121199
Pulled By: pdillinger
fbshipit-source-id: 8f21732c3d8909777e3240e4ac3123d73140326a
Summary:
The first version of filter_bench has selectable key size
but that size does not vary throughout a test run. This artificially
favors "branchy" hash functions like the existing BloomHash,
MurmurHash1, probably because of optimal return for branch prediction.
This change primarily varies those key sizes from -2 to +2 bytes vs.
the average selected size. We also set the default key size at 24 to
better reflect our best guess of typical key size.
But steadily random key sizes may not be realistic either. So this
change introduces a new filter_bench option:
-vary_key_size_log2_interval=n where the same key size is used 2^n
times and then changes to another size. I've set the default at 5
(32 times same size) as a compromise between deployments with
rather consistent vs. rather variable key sizes. On my Skylake
system, the performance boost to MurmurHash1 largely lies between
n=10 and n=15.
Also added -vary_key_alignment (bool, now default=true), though this
doesn't currently seem to matter in hash functions under
consideration.
This change also does a "dry run" for each testing scenario, to improve
the accuracy of those numbers, as there was more difference between
scenarios than expected. Subtracting gross test run times from dry run
times is now also embedded in the output, because these "net" times are
generally the most useful.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5933
Differential Revision: D18121683
Pulled By: pdillinger
fbshipit-source-id: 3c7efee1c5661a5fe43de555e786754ddf80dc1e
Summary:
This is an internal, file-local "feature" that is not used and
potentially confusing.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5961
Test Plan: make check
Differential Revision: D18099018
Pulled By: pdillinger
fbshipit-source-id: 7870627eeed09941d12538ec55d10d2e164fc716
Summary:
Amongst other things, PR https://github.com/facebook/rocksdb/issues/5504 refactored the filter block readers so that
only the filter block contents are stored in the block cache (as opposed to the
earlier design where the cache stored the filter block reader itself, leading to
potentially dangling pointers and concurrency bugs). However, this change
introduced a performance hit since with the new code, the metadata fields are
re-parsed upon every access. This patch reunites the block contents with the
filter bits reader to eliminate this overhead; since this is still a self-contained
pure data object, it is safe to store it in the cache. (Note: this is similar to how
the zstd digest is handled.)
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5936
Test Plan:
make asan_check
filter_bench results for the old code:
```
$ ./filter_bench -quick
WARNING: Assertions are enabled; benchmarks unnecessarily slow
Building...
Build avg ns/key: 26.7153
Number of filters: 16669
Total memory (MB): 200.009
Bits/key actual: 10.0647
----------------------------
Inside queries...
Dry run (46b) ns/op: 33.4258
Single filter ns/op: 42.5974
Random filter ns/op: 217.861
----------------------------
Outside queries...
Dry run (25d) ns/op: 32.4217
Single filter ns/op: 50.9855
Random filter ns/op: 219.167
Average FP rate %: 1.13993
----------------------------
Done. (For more info, run with -legend or -help.)
$ ./filter_bench -quick -use_full_block_reader
WARNING: Assertions are enabled; benchmarks unnecessarily slow
Building...
Build avg ns/key: 26.5172
Number of filters: 16669
Total memory (MB): 200.009
Bits/key actual: 10.0647
----------------------------
Inside queries...
Dry run (46b) ns/op: 32.3556
Single filter ns/op: 83.2239
Random filter ns/op: 370.676
----------------------------
Outside queries...
Dry run (25d) ns/op: 32.2265
Single filter ns/op: 93.5651
Random filter ns/op: 408.393
Average FP rate %: 1.13993
----------------------------
Done. (For more info, run with -legend or -help.)
```
With the new code:
```
$ ./filter_bench -quick
WARNING: Assertions are enabled; benchmarks unnecessarily slow
Building...
Build avg ns/key: 25.4285
Number of filters: 16669
Total memory (MB): 200.009
Bits/key actual: 10.0647
----------------------------
Inside queries...
Dry run (46b) ns/op: 31.0594
Single filter ns/op: 43.8974
Random filter ns/op: 226.075
----------------------------
Outside queries...
Dry run (25d) ns/op: 31.0295
Single filter ns/op: 50.3824
Random filter ns/op: 226.805
Average FP rate %: 1.13993
----------------------------
Done. (For more info, run with -legend or -help.)
$ ./filter_bench -quick -use_full_block_reader
WARNING: Assertions are enabled; benchmarks unnecessarily slow
Building...
Build avg ns/key: 26.5308
Number of filters: 16669
Total memory (MB): 200.009
Bits/key actual: 10.0647
----------------------------
Inside queries...
Dry run (46b) ns/op: 33.2968
Single filter ns/op: 58.6163
Random filter ns/op: 291.434
----------------------------
Outside queries...
Dry run (25d) ns/op: 32.1839
Single filter ns/op: 66.9039
Random filter ns/op: 292.828
Average FP rate %: 1.13993
----------------------------
Done. (For more info, run with -legend or -help.)
```
Differential Revision: D17991712
Pulled By: ltamasi
fbshipit-source-id: 7ea205550217bfaaa1d5158ebd658e5832e60f29
Summary:
FullFilterBitsReader, after creating in BloomFilterPolicy, was
responsible for decoding metadata bits. This meant that
FullFilterBitsReader::MayMatch had some metadata checks in order to
implement "always true" or "always false" functionality in the case
of inconsistent or trivial metadata. This made for ugly
mixing-of-concerns code and probably had some runtime cost. It also
didn't really support plugging in alternative filter implementations
with extensions to the existing metadata schema.
BloomFilterPolicy::GetFilterBitsReader is now (exclusively) responsible
for decoding filter metadata bits and constructing appropriate instances
deriving from FilterBitsReader. "Always false" and "always true" derived
classes allow FullFilterBitsReader not to be concerned with handling of
trivial or inconsistent metadata. This also makes for easy expansion
to alternative filter implementations in new, alternative derived
classes. This change makes calls to FilterBitsReader::MayMatch
*necessarily* virtual because there's now more than one built-in
implementation. Compared with the previous implementation's extra
'if' checks in MayMatch, there's no consistent performance difference,
measured by (an older revision of) filter_bench (differences here seem
to be within noise):
Inside queries...
- Dry run (407) ns/op: 35.9996
+ Dry run (407) ns/op: 35.2034
- Single filter ns/op: 47.5483
+ Single filter ns/op: 47.4034
- Batched, prepared ns/op: 43.1559
+ Batched, prepared ns/op: 42.2923
...
- Random filter ns/op: 150.697
+ Random filter ns/op: 149.403
----------------------------
Outside queries...
- Dry run (980) ns/op: 34.6114
+ Dry run (980) ns/op: 34.0405
- Single filter ns/op: 56.8326
+ Single filter ns/op: 55.8414
- Batched, prepared ns/op: 48.2346
+ Batched, prepared ns/op: 47.5667
- Random filter ns/op: 155.377
+ Random filter ns/op: 153.942
Average FP rate %: 1.1386
Also, the FullFilterBitsReader ctor was responsible for a surprising
amount of CPU in production, due in part to inefficient determination of
the CACHE_LINE_SIZE used to construct the filter being read. The
overwhelming common case (same as my CACHE_LINE_SIZE) is now
substantially optimized, as shown with filter_bench with
-new_reader_every=1 (old option - see below) (repeatable result):
Inside queries...
- Dry run (453) ns/op: 118.799
+ Dry run (453) ns/op: 105.869
- Single filter ns/op: 82.5831
+ Single filter ns/op: 74.2509
...
- Random filter ns/op: 224.936
+ Random filter ns/op: 194.833
----------------------------
Outside queries...
- Dry run (aa1) ns/op: 118.503
+ Dry run (aa1) ns/op: 104.925
- Single filter ns/op: 90.3023
+ Single filter ns/op: 83.425
...
- Random filter ns/op: 220.455
+ Random filter ns/op: 175.7
Average FP rate %: 1.13886
However PR#5936 has/will reclaim most of this cost. After that PR, the optimization of this code path is likely negligible, but nonetheless it's clear we aren't making performance any worse.
Also fixed inadequate check of consistency between filter data size and
num_lines. (Unit test updated.)
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5941
Test Plan:
previously added unit tests FullBloomTest.CorruptFilters and
FullBloomTest.RawSchema
Differential Revision: D18018353
Pulled By: pdillinger
fbshipit-source-id: 8e04c2b4a7d93223f49a237fd52ef2483929ed9c
Summary:
RocksDB has a MultiGet() API that implements batched key lookup for higher performance (https://github.com/facebook/rocksdb/blob/master/include/rocksdb/db.h#L468). Currently, batching is implemented in BlockBasedTableReader::MultiGet() for SST file lookups. One of the ways it improves performance is by pipelining bloom filter lookups (by prefetching required cachelines for all the keys in the batch, and then doing the probe) and thus hiding the cache miss latency. The same concept can be extended to the memtable as well. This PR involves implementing a pipelined bloom filter lookup in DynamicBloom, and implementing MemTable::MultiGet() that can leverage it.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5818
Test Plan:
Existing tests
Performance Test:
Ran the below command which fills up the memtable and makes sure there are no flushes and then call multiget. Ran it on master and on the new change and see atleast 1% performance improvement across all the test runs I did. Sometimes the improvement was upto 5%.
TEST_TMPDIR=/data/users/$USER/benchmarks/feature/ numactl -C 10 ./db_bench -benchmarks="fillseq,multireadrandom" -num=600000 -compression_type="none" -level_compaction_dynamic_level_bytes -write_buffer_size=200000000 -target_file_size_base=200000000 -max_bytes_for_level_base=16777216 -reads=90000 -threads=1 -compression_type=none -cache_size=4194304000 -batch_size=32 -disable_auto_compactions=true -bloom_bits=10 -cache_index_and_filter_blocks=true -pin_l0_filter_and_index_blocks_in_cache=true -multiread_batched=true -multiread_stride=4 -statistics -memtable_whole_key_filtering=true -memtable_bloom_size_ratio=10
Differential Revision: D17578869
Pulled By: vjnadimpalli
fbshipit-source-id: 23dc651d9bf49db11d22375bf435708875a1f192
Summary:
Fixed some spots where converting size_t or uint_fast32_t to
uint32_t. Wrapped mt19937 in a new Random32 class to avoid future
such traps.
NB: I tried using Random32::Uniform (std::uniform_int_distribution) in
filter_bench instead of fastrange, but that more than doubled the dry
run time! So I added fastrange as Random32::Uniformish. ;)
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5894
Test Plan: USE_CLANG=1 build, and manual re-run filter_bench
Differential Revision: D17825131
Pulled By: pdillinger
fbshipit-source-id: 68feee333b5f8193c084ded760e3d6679b405ecd
Summary:
Example: using the tool before and after PR https://github.com/facebook/rocksdb/issues/5784 shows that
the refactoring, presumed performance-neutral, actually sped up SST
filters by about 3% to 8% (repeatable result):
Before:
- Dry run ns/op: 22.4725
- Single filter ns/op: 51.1078
- Random filter ns/op: 120.133
After:
+ Dry run ns/op: 22.2301
+ Single filter run ns/op: 47.4313
+ Random filter ns/op: 115.9
Only tests filters for the block-based table (full filters and
partitioned filters - same implementation; not block-based filters),
which seems to be the recommended format/implementation.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5825
Differential Revision: D17804987
Pulled By: pdillinger
fbshipit-source-id: 0f18a9c254c57f7866030d03e7fa4ba503bac3c5
Summary:
Broken type for shift in PR#5834. Fixing code means fixing
expected values in test.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5882
Test Plan: thisisthetest
Differential Revision: D17746136
Pulled By: pdillinger
fbshipit-source-id: d3c456ed30b433d55fcab6fc7d836940fe3b46b8
Summary:
There was significant untested logic in FullFilterBitsReader in
the handling of serialized Bloom filter bits that cannot be generated by
FullFilterBitsBuilder in the current compilation. These now test many of
those corner-case behaviors, including bad metadata or filters created
with different cache line size than the current compiled-in value.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5834
Test Plan: thisisthetest
Differential Revision: D17726372
Pulled By: pdillinger
fbshipit-source-id: fb7b8003b5a8e6fb4666fe95206128f3d5835fc7
Summary:
Further apply formatter to more recent commits.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5830
Test Plan: Run all existing tests.
Differential Revision: D17488031
fbshipit-source-id: 137458fd94d56dd271b8b40c522b03036943a2ab
Summary:
Some recent commits might not have passed through the formatter. I formatted recent 45 commits. The script hangs for more commits so I stopped there.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5827
Test Plan: Run all existing tests.
Differential Revision: D17483727
fbshipit-source-id: af23113ee63015d8a43d89a3bc2c1056189afe8f
Summary:
clang-analyzer has uncovered a bunch of places where the code is relying
on pointers being valid and one case (in VectorIterator) where a moved-from
object is being used:
In file included from db/range_tombstone_fragmenter.cc:17:
./util/vector_iterator.h:23:18: warning: Method called on moved-from object 'keys' of type 'std::vector'
current_(keys.size()) {
^~~~~~~~~~~
1 warning generated.
utilities/persistent_cache/block_cache_tier_file.cc:39:14: warning: Called C++ object pointer is null
Status s = env->NewRandomAccessFile(filepath, file, opt);
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
utilities/persistent_cache/block_cache_tier_file.cc:47:19: warning: Called C++ object pointer is null
Status status = env_->GetFileSize(Path(), size);
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
utilities/persistent_cache/block_cache_tier_file.cc:290:14: warning: Called C++ object pointer is null
Status s = env_->FileExists(Path());
^~~~~~~~~~~~~~~~~~~~~~~~
utilities/persistent_cache/block_cache_tier_file.cc:363:35: warning: Called C++ object pointer is null
CacheWriteBuffer* const buf = alloc_->Allocate();
^~~~~~~~~~~~~~~~~~
utilities/persistent_cache/block_cache_tier_file.cc:399:41: warning: Called C++ object pointer is null
const uint64_t file_off = buf_doff_ * alloc_->BufferSize();
^~~~~~~~~~~~~~~~~~~~
utilities/persistent_cache/block_cache_tier_file.cc:463:33: warning: Called C++ object pointer is null
size_t start_idx = lba.off_ / alloc_->BufferSize();
^~~~~~~~~~~~~~~~~~~~
utilities/persistent_cache/block_cache_tier_file.cc:515:5: warning: Called C++ object pointer is null
alloc_->Deallocate(bufs_[i]);
^~~~~~~~~~~~~~~~~~~~~~~~~~~~
7 warnings generated.
ar: creating librocksdb_debug.a
utilities/memory/memory_test.cc:68:25: warning: Called C++ object pointer is null
cache_set->insert(db->GetDBOptions().row_cache.get());
^~~~~~~~~~~~~~~~~~
1 warning generated.
The patch fixes these by adding assertions and explicitly passing in zero
when initializing VectorIterator::current_ (which preserves the existing
behavior).
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5821
Test Plan: Ran make check and make analyze to make sure the warnings have disappeared.
Differential Revision: D17455949
Pulled By: ltamasi
fbshipit-source-id: 363619618ea649a0674287f9f3b3393e390571ee
Summary:
Manual compaction may bring in very high load because sometime the amount of data involved in a compaction could be large, which may affect online service. So it would be good if the running compaction making the server busy can be stopped immediately. In this implementation, stopping manual compaction condition is only checked in slow process. We let deletion compaction and trivial move go through.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/3971
Test Plan: add tests at more spots.
Differential Revision: D17369043
fbshipit-source-id: 575a624fb992ce0bb07d9443eb209e547740043c
Summary:
Refactoring to consolidate implementation details of legacy
Bloom filters. This helps to organize and document some related,
obscure code.
Also added make/cpp var TEST_CACHE_LINE_SIZE so that it's easy to
compile and run unit tests for non-native cache line size. (Fixed a
related test failure in db_properties_test.)
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5784
Test Plan:
make check, including Recently added Bloom schema unit tests
(in ./plain_table_db_test && ./bloom_test), and including with
TEST_CACHE_LINE_SIZE=128U and TEST_CACHE_LINE_SIZE=256U. Tested the
schema tests with temporary fault injection into new implementations.
Some performance testing with modified unit tests suggest a small to moderate
improvement in speed.
Differential Revision: D17381384
Pulled By: pdillinger
fbshipit-source-id: ee42586da996798910fc45ac0b6289147f16d8df
Summary:
This will allow us to fix history by having the code changes for PR#5784 properly attributed to it.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5810
Differential Revision: D17400231
Pulled By: pdillinger
fbshipit-source-id: 2da8b1cdf2533cfedb35b5526eadefb38c291f09
Summary:
file_reader_writer.h and .cc contain several files and helper function, and it's hard to navigate. Separate it to multiple files and put them under file/
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5803
Test Plan: Build whole project using make and cmake.
Differential Revision: D17374550
fbshipit-source-id: 10efca907721e7a78ed25bbf74dc5410dea05987
Summary:
DynamicBloom unit test now tests non-sequential as well as
sequential keys in testing FP rates. Also now verifies larger structures.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5805
Test Plan: thisisthetest
Differential Revision: D17398109
Pulled By: pdillinger
fbshipit-source-id: 374074206c76d242efa378afc27830448a0e892a
Summary:
prefetch data for following block，avoid cache miss when doing crc caculate
I do performance test at kunpeng-920 server(arm-v8, 64core@2.6GHz)
./db_bench --benchmarks=crc32c --block_size=500000000
before optimise : 587313.500 micros/op 1 ops/sec; 811.9 MB/s (500000000 per op)
after optimise : 289248.500 micros/op 3 ops/sec; 1648.5 MB/s (500000000 per op)
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5773
Differential Revision: D17347339
fbshipit-source-id: bfcd74f0f0eb4b322b959be68019ddcaae1e3341
Summary:
Use delete to disable automatic generated methods instead of private, and put the constructor together for more clear.This modification cause the unused field warning, so add unused attribute to disable this warning.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5009
Differential Revision: D17288733
fbshipit-source-id: 8a767ce096f185f1db01bd28fc88fef1cdd921f3
Summary:
Bug found by valgrind. New DynamicBloom wasn't allocating in
block sizes. New assertion added that probes starting in final word
would be in bounds.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5783
Test Plan: ROCKSDB_VALGRIND_RUN=1 DISABLE_JEMALLOC=1 valgrind --leak-check=full ./dynamic_bloom_test
Differential Revision: D17270623
Pulled By: pdillinger
fbshipit-source-id: 1e0407504b875133a771383cd488c70f91be2b87
Summary:
Check that we don't accidentally change the on-disk format of
existing Bloom filter implementations, including for various
CACHE_LINE_SIZE (by changing temporarily).
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5778
Test Plan: thisisthetest
Differential Revision: D17269630
Pulled By: pdillinger
fbshipit-source-id: c77017662f010a77603b7d475892b1f0d5563d8b
Summary:
FullFilterBitsBuilder::CalculateSpace use CACHE_LINE_SIZE which is 64@X86 but 128@ARM64
when it run bloom_test.FullVaryingLengths it failed on ARM64 server,
the assert can be fixed by change 128->CACHE_LINE_SIZE*2 as merged
ASSERT_LE(FilterSize(), (size_t)((length * 10 / 8) + CACHE_LINE_SIZE * 2 + 5)) << length;
run bloom_test
before fix:
/root/rocksdb-master/util/bloom_test.cc:281: Failure
Expected: (FilterSize()) <= ((size_t)((length * 10 / 8) + 128 + 5)), actual: 389 vs 383
200
[ FAILED ] FullBloomTest.FullVaryingLengths (32 ms)
[----------] 4 tests from FullBloomTest (32 ms total)
[----------] Global test environment tear-down
[==========] 7 tests from 2 test cases ran. (116 ms total)
[ PASSED ] 6 tests.
[ FAILED ] 1 test, listed below:
[ FAILED ] FullBloomTest.FullVaryingLengths
after fix:
Filters: 37 good, 0 mediocre
[ OK ] FullBloomTest.FullVaryingLengths (90 ms)
[----------] 4 tests from FullBloomTest (90 ms total)
[----------] Global test environment tear-down
[==========] 7 tests from 2 test cases ran. (174 ms total)
[ PASSED ] 7 tests.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5745
Differential Revision: D17076047
fbshipit-source-id: e7beb5d55d4855fceb2b84bc8119a6b0759de635
Summary:
Since DynamicBloom is now only used in-memory, we're free to
change it without schema compatibility issues. The new implementation
is drawn from (with manifest permission)
303542a767/bloom_simulation_tests/foo.cc (L613)
This has several speed advantages over the prior implementation:
* Uses fastrange instead of %
* Minimum logic to determine first (and all) probed memory addresses
* (Major) Two probes per 64-bit memory fetch/write.
* Very fast and effective (murmur-like) hash expansion/re-mixing. (At
least on recent CPUs, integer multiplication is very cheap.)
While a Bloom filter with 512-bit cache locality has about a 1.15x FP
rate penalty (e.g. 0.84% to 0.97%), further restricting to two probes
per 64 bits incurs an additional 1.12x FP rate penalty (e.g. 0.97% to
1.09%). Nevertheless, the unit tests show no "mediocre" FP rate samples,
unlike the old implementation with more erratic FP rates.
Especially for the memtable, we expect speed to outweigh somewhat higher
FP rates. For example, a negative table query would have to be 1000x
slower than a BF query to justify doubling BF query time to shave 10% off
FP rate (working assumption around 1% FP rate). While that seems likely
for SSTs, my data suggests a speed factor of roughly 50x for the memtable
(vs. BF; ~1.5% lower write throughput when enabling memtable Bloom
filter, after this change). Thus, it's probably not worth even 5% more
time in the Bloom filter to shave off 1/10th of the Bloom FP rate, or 0.1%
in absolute terms, and it's probably at least 20% slower to recoup that
much FP rate from this new implementation. Because of this, we do not see
a need for a 'locality' option that affects the MemTable Bloom filter
and have decoupled the MemTable Bloom filter from Options::bloom_locality.
Note that just 3% more memory to the Bloom filter (10.3 bits per key vs.
just 10) is able to make up for the ~12% FP rate drop in the new
implementation:
[] # Nearly "ideal" FP-wise but reasonably fast cache-local implementation
[~/wormhashing/bloom_simulation_tests] ./foo_gcc_IMPL_CACHE_WORM64_FROM32_any.out 10000000 6 10 $RANDOM 100000000
./foo_gcc_IMPL_CACHE_WORM64_FROM32_any.out time: 3.29372 sampled_fp_rate: 0.00985956 ...
[] # Close match to this new implementation
[~/wormhashing/bloom_simulation_tests] ./foo_gcc_IMPL_CACHE_MUL64_BLOCK_FROM32_any.out 10000000 6 10.3 $RANDOM 100000000
./foo_gcc_IMPL_CACHE_MUL64_BLOCK_FROM32_any.out time: 2.10072 sampled_fp_rate: 0.00985655 ...
[] # Old locality=1 implementation
[~/wormhashing/bloom_simulation_tests] ./foo_gcc_IMPL_CACHE_ROCKSDB_DYNAMIC_any.out 10000000 6 10 $RANDOM 100000000
./foo_gcc_IMPL_CACHE_ROCKSDB_DYNAMIC_any.out time: 3.95472 sampled_fp_rate: 0.00988943 ...
Also note the dramatic speed improvement vs. alternatives.
--
Performance unit test: DynamicBloomTest.concurrent_with_perf is updated
to report more precise timing data. (Measure running time of each
thread, not just longest running thread, etc.) Results averaged over
various sizes enabled with --enable_perf and 20 runs each; old dynamic
bloom refers to locality=1, the faster of the old:
old dynamic bloom, avg add latency = 65.6468
new dynamic bloom, avg add latency = 44.3809
old dynamic bloom, avg query latency = 50.6485
new dynamic bloom, avg query latency = 43.2186
old avg parallel add latency = 41.678
new avg parallel add latency = 24.5238
old avg parallel hit latency = 14.6322
new avg parallel hit latency = 12.3939
old avg parallel miss latency = 16.7289
new avg parallel miss latency = 12.2134
Tested on a dedicated 64-bit production machine at Facebook. Significant
improvement all around.
Despite now using std::atomic<uint64_t>, quick before-and-after test on
a 32-bit machine (Intel Atom N270, released 2008) shows no regression in
performance, in some cases modest improvement.
--
Performance integration test (synthetic): with DEBUG_LEVEL=0, used
TEST_TMPDIR=/dev/shm ./db_bench --benchmarks=fillrandom,readmissing,readrandom,stats --num=2000000
and optionally with -memtable_whole_key_filtering -memtable_bloom_size_ratio=0.01
300 runs each configuration.
Write throughput change by enabling memtable bloom:
Old locality=0: -3.06%
Old locality=1: -2.37%
New: -1.50%
conclusion -> seems to substantially close the gap
Readmissing throughput change by enabling memtable bloom:
Old locality=0: +34.47%
Old locality=1: +34.80%
New: +33.25%
conclusion -> maybe a small new penalty from FP rate
Readrandom throughput change by enabling memtable bloom:
Old locality=0: +31.54%
Old locality=1: +31.13%
New: +30.60%
conclusion -> maybe also from FP rate (after memtable flush)
--
Another conclusion we can draw from this new implementation is that the
existing 32-bit hash function is not inherently crippling the Bloom
filter speed or accuracy, below about 5 million keys. For speed, the
implementation is essentially the same whether starting with 32-bits or
64-bits of hash; it just determines whether the first multiplication
after fastrange is a pseudorandom expansion or needed re-mix. Note that
this multiplication can occur while memory is fetching.
For accuracy, in a standard configuration, you need about 5 million
keys before you have about a 1.1x FP penalty due to using a
32-bit hash vs. 64-bit:
[~/wormhashing/bloom_simulation_tests] ./foo_gcc_IMPL_CACHE_MUL64_BLOCK_FROM32_any.out $((5 * 1000 * 1000 * 10)) 6 10 $RANDOM 100000000
./foo_gcc_IMPL_CACHE_MUL64_BLOCK_FROM32_any.out time: 2.52069 sampled_fp_rate: 0.0118267 ...
[~/wormhashing/bloom_simulation_tests] ./foo_gcc_IMPL_CACHE_MUL64_BLOCK_any.out $((5 * 1000 * 1000 * 10)) 6 10 $RANDOM 100000000
./foo_gcc_IMPL_CACHE_MUL64_BLOCK_any.out time: 2.43871 sampled_fp_rate: 0.0109059
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5762
Differential Revision: D17214194
Pulled By: pdillinger
fbshipit-source-id: ad9da031772e985fd6b62a0e1db8e81892520595
Summary:
DynamicBloom was being used both for memory-only and for on-disk filters, as part of the PlainTable format. To set up enhancements to the memtable Bloom filter, this splits the code into two copies and removes unused features from each copy. Adds test PlainTableDBTest.BloomSchema to ensure no accidental change to that format.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5767
Differential Revision: D17206963
Pulled By: pdillinger
fbshipit-source-id: 6cce8d55305ed0df051b4c58bdc98c8ad81d0553
Summary:
MyRocks currently sets `max_write_buffer_number_to_maintain` in order to maintain enough history for transaction conflict checking. The effectiveness of this approach depends on the size of memtables. When memtables are small, it may not keep enough history; when memtables are large, this may consume too much memory.
We are proposing a new way to configure memtable list history: by limiting the memory usage of immutable memtables. The new option is `max_write_buffer_size_to_maintain` and it will take precedence over the old `max_write_buffer_number_to_maintain` if they are both set to non-zero values. The new option accounts for the total memory usage of flushed immutable memtables and mutable memtable. When the total usage exceeds the limit, RocksDB may start dropping immutable memtables (which is also called trimming history), starting from the oldest one.
The semantics of the old option actually works both as an upper bound and lower bound. History trimming will start if number of immutable memtables exceeds the limit, but it will never go below (limit-1) due to history trimming.
In order the mimic the behavior with the new option, history trimming will stop if dropping the next immutable memtable causes the total memory usage go below the size limit. For example, assuming the size limit is set to 64MB, and there are 3 immutable memtables with sizes of 20, 30, 30. Although the total memory usage is 80MB > 64MB, dropping the oldest memtable will reduce the memory usage to 60MB < 64MB, so in this case no memtable will be dropped.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5022
Differential Revision: D14394062
Pulled By: miasantreble
fbshipit-source-id: 60457a509c6af89d0993f988c9b5c2aa9e45f5c5
Summary:
Crc32c Parallel computation coding optimization:
Macro unfolding removes the "for" loop and is good to decrease branch-miss in arm64 micro architecture
1024 Bytes is divided into 8(head) + 1008( 6 * 7 * 3 * 8 ) + 8(tail) three parts
Macro unfolding 42 loops to 6 CRC32C7X24BYTESs
1 CRC32C7X24BYTES containing 7 CRC32C24BYTESs
1, crc32c_test
[==========] Running 4 tests from 1 test case.
[----------] Global test environment set-up.
[----------] 4 tests from CRC
[ RUN ] CRC.StandardResults
[ OK ] CRC.StandardResults (1 ms)
[ RUN ] CRC.Values
[ OK ] CRC.Values (0 ms)
[ RUN ] CRC.Extend
[ OK ] CRC.Extend (0 ms)
[ RUN ] CRC.Mask
[ OK ] CRC.Mask (0 ms)
[----------] 4 tests from CRC (1 ms total)
[----------] Global test environment tear-down
[==========] 4 tests from 1 test case ran. (1 ms total)
[ PASSED ] 4 tests.
2, db_bench --benchmarks="crc32c"
crc32c : 0.218 micros/op 4595390 ops/sec; 17950.7 MB/s (4096 per op)
3, repeated crc32c_test case 60000 times
perf stat -e branch-miss -- ./crc32c_test
before optimization:
739,426,504 branch-miss
after optimization:
1,128,572 branch-miss
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5675
Differential Revision: D16989210
fbshipit-source-id: 7204e6069bb6ed066d49c2d1b3ac385065a98557
Summary:
PR https://github.com/facebook/rocksdb/issues/5584 decoupled the uncompression dictionary object from the underlying block data; however, this defeats the purpose of the digested ZSTD dictionary, since the whole point
of the digest is to create it once and reuse it over and over again. This patch goes back to
storing the uncompression dictionary itself in the cache (which should be now safe to do,
since it no longer includes a Statistics pointer), while preserving the rest of the refactoring.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5645
Test Plan: make asan_check
Differential Revision: D16551864
Pulled By: ltamasi
fbshipit-source-id: 2a7e2d34bb16e70e3c816506d5afe1d842057800
Summary:
there is no need to return void*, as
std:🧵:thread(Func&& f, Args&&... args ) only requires `Func` to
be callable.
Signed-off-by: Kefu Chai <tchaikov@gmail.com>
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5709
Differential Revision: D16832894
fbshipit-source-id: a1e1b876fa8d55589ef5feb5b27f3a435068b747
Summary:
In previous https://github.com/facebook/rocksdb/issues/5079, we added user-specified timestamp to `DB::Get()` and `DB::Put()`. Limitation is that these two functions may cause extra memory allocation and key copy. The reason is that `WriteBatch` does not allocate extra memory for timestamps because it is not aware of timestamp size, and we did not provide an API to assign/update timestamp of each key within a `WriteBatch`.
We address these issues in this PR by doing the following.
1. Add a `timestamp_size_` to `WriteBatch` so that `WriteBatch` can take timestamps into account when calling `WriteBatch::Put`, `WriteBatch::Delete`, etc.
2. Add APIs `WriteBatch::AssignTimestamp` and `WriteBatch::AssignTimestamps` so that application can assign/update timestamps for each key in a `WriteBatch`.
3. Avoid key copy in `GetImpl` by adding new constructor to `LookupKey`.
Test plan (on devserver):
```
$make clean && COMPILE_WITH_ASAN=1 make -j32 all
$./db_basic_test --gtest_filter=Timestamp/DBBasicTestWithTimestampWithParam.PutAndGet/*
$make check
```
If the API extension looks good, I will add more unit tests.
Some simple benchmark using db_bench.
```
$rm -rf /dev/shm/dbbench/* && TEST_TMPDIR=/dev/shm ./db_bench -benchmarks=fillseq,readrandom -num=1000000
$rm -rf /dev/shm/dbbench/* && TEST_TMPDIR=/dev/shm ./db_bench -benchmarks=fillrandom -num=1000000 -disable_wal=true
```
Master is at a78503bd6c.
```
| | readrandom | fillrandom |
| master | 15.53 MB/s | 25.97 MB/s |
| PR5502 | 16.70 MB/s | 25.80 MB/s |
```
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5502
Differential Revision: D16340894
Pulled By: riversand963
fbshipit-source-id: 51132cf792be07d1efc3ac33f5768c4ee2608bb8
Summary:
RocksDB has historically stored uncompression dictionary objects in the block
cache as opposed to storing just the block contents. This neccesitated
evicting the object upon table close. With the new code, only the raw blocks
are stored in the cache, eliminating the need for eviction.
In addition, the patch makes the following improvements:
1) Compression dictionary blocks are now prefetched/pinned similarly to
index/filter blocks.
2) A copy operation got eliminated when the uncompression dictionary is
retrieved.
3) Errors related to retrieving the uncompression dictionary are propagated as
opposed to silently ignored.
Note: the patch temporarily breaks the compression dictionary evicition stats.
They will be fixed in a separate phase.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5584
Test Plan: make asan_check
Differential Revision: D16344151
Pulled By: ltamasi
fbshipit-source-id: 2962b295f5b19628f9da88a3fcebbce5a5017a7b
Summary:
Fixing a corner case crash when there was no data read from file, but status is still OK
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5586
Differential Revision: D16348117
Pulled By: elipoz
fbshipit-source-id: f97973308024f020d8be79ca3c56466b84d80656
Summary:
Crc32c Parallel computation optimization:
Algorithm comes from Intel whitepaper: [crc-iscsi-polynomial-crc32-instruction-paper](https://www.intel.com/content/dam/www/public/us/en/documents/white-papers/crc-iscsi-polynomial-crc32-instruction-paper.pdf)
Input data is divided into three equal-sized blocks
Three parallel blocks (crc0, crc1, crc2) for 1024 Bytes
One Block: 42(BLK_LENGTH) * 8(step length: crc32c_u64) bytes
1. crc32c_test:
```
[==========] Running 4 tests from 1 test case.
[----------] Global test environment set-up.
[----------] 4 tests from CRC
[ RUN ] CRC.StandardResults
[ OK ] CRC.StandardResults (1 ms)
[ RUN ] CRC.Values
[ OK ] CRC.Values (0 ms)
[ RUN ] CRC.Extend
[ OK ] CRC.Extend (0 ms)
[ RUN ] CRC.Mask
[ OK ] CRC.Mask (0 ms)
[----------] 4 tests from CRC (1 ms total)
[----------] Global test environment tear-down
[==========] 4 tests from 1 test case ran. (1 ms total)
[ PASSED ] 4 tests.
```
2. RocksDB benchmark: db_bench --benchmarks="crc32c"
```
Linear Arm crc32c:
crc32c: 1.005 micros/op 995133 ops/sec; 3887.2 MB/s (4096 per op)
```
```
Parallel optimization with Armv8 crypto extension:
crc32c: 0.419 micros/op 2385078 ops/sec; 9316.7 MB/s (4096 per op)
```
It gets ~2.4x speedup compared to linear Arm crc32c instructions.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5494
Differential Revision: D16340806
fbshipit-source-id: 95dae9a5b646fd20a8303671d82f17b2e162e945
Summary:
Added support for sequential read-ahead file that can prefetch the read data and later serve it from internal cache buffer.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5580
Differential Revision: D16287082
Pulled By: elipoz
fbshipit-source-id: a3e7ad9643d377d39352ff63058ce050ec31dcf3
Summary:
RandomAccessFileReader.for_compaction_ doesn't seem to be used anymore. Remove it.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5572
Test Plan: USE_CLANG=1 make all check -j
Differential Revision: D16286178
fbshipit-source-id: aa338049761033dfbe5e8b1707bbb0be2df5be7e
Summary:
When 'HAVE_ARM64_CRC' is set, the blew methods:
- bool rocksdb::crc32c::isSSE42()
- bool rocksdb::crc32c::isPCLMULQDQ()
are defined but not used, the unused-function is raised
when do rocksdb build.
This patch try to cleanup these warnings by add ifndef,
if it build under the HAVE_ARM64_CRC, we will not define
`isSSE42` and `isPCLMULQDQ`.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5565
Differential Revision: D16233654
fbshipit-source-id: c32a9dda7465dbf65f9ccafef159124db92cdffd
Summary:
Current PosixLogger performs IO operations using posix calls. Thus the
current implementation will not work for non-posix env. Created a new
logger class EnvLogger that uses env specific WritableFileWriter for IO operations.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5491
Test Plan: make check
Differential Revision: D15909002
Pulled By: ggaurav28
fbshipit-source-id: 13a8105176e8e42db0c59798d48cb6a0dbccc965
Summary:
Sometimes it is helpful to fetch the whole history of stats after benchmark runs. Add such an option
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5532
Test Plan: Run the benchmark manually and observe the output is as expected.
Differential Revision: D16097764
fbshipit-source-id: 10b5b735a22a18be198b8f348be11f11f8806904
Summary:
Enhancement to MultiGet batching to read data blocks required for keys in a batch in parallel from disk. It uses Env::MultiRead() API to read multiple blocks and reduce latency.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5464
Test Plan:
1. make check
2. make asan_check
3. make asan_crash
Differential Revision: D15911771
Pulled By: anand1976
fbshipit-source-id: 605036b9af0f90ca0020dc87c3a86b4da6e83394
Summary:
The first key is used to defer reading the data block until this file gets to the top of merging iterator's heap. For short range scans, most files never make it to the top of the heap, so this change can reduce read amplification by a lot sometimes.
Consider the following workload. There are a few data streams (we'll be calling them "logs"), each stream consisting of a sequence of blobs (we'll be calling them "records"). Each record is identified by log ID and a sequence number within the log. RocksDB key is concatenation of log ID and sequence number (big endian). Reads are mostly relatively short range scans, each within a single log. Writes are mostly sequential for each log, but writes to different logs are randomly interleaved. Compactions are disabled; instead, when we accumulate a few tens of sst files, we create a new column family and start writing to it.
So, a typical sst file consists of a few ranges of blocks, each range corresponding to one log ID (we use FlushBlockPolicy to cut blocks at log boundaries). A typical read would go like this. First, iterator Seek() reads one block from each sst file. Then a series of Next()s move through one sst file (since writes to each log are mostly sequential) until the subiterator reaches the end of this log in this sst file; then Next() switches to the next sst file and reads sequentially from that, and so on. Often a range scan will only return records from a small number of blocks in small number of sst files; in this case, the cost of initial Seek() reading one block from each file may be bigger than the cost of reading the actually useful blocks.
Neither iterate_upper_bound nor bloom filters can prevent reading one block from each file in Seek(). But this PR can: if the index contains first key from each block, we don't have to read the block until this block actually makes it to the top of merging iterator's heap, so for short range scans we won't read any blocks from most of the sst files.
This PR does the deferred block loading inside value() call. This is not ideal: there's no good way to report an IO error from inside value(). As discussed with siying offline, it would probably be better to change InternalIterator's interface to explicitly fetch deferred value and get status. I'll do it in a separate PR.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5289
Differential Revision: D15256423
Pulled By: al13n321
fbshipit-source-id: 750e4c39ce88e8d41662f701cf6275d9388ba46a
Summary:
Currently the read-ahead logic for user reads and compaction reads go through different code paths where compaction reads create new table readers and use `ReadaheadRandomAccessFile`. This change is to unify read-ahead logic to use read-ahead in BlockBasedTableReader::InitDataBlock(). As a result of the change `ReadAheadRandomAccessFile` class and `new_table_reader_for_compaction_inputs` option will no longer be used.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5431
Test Plan:
make check
Here is the benchmarking - https://gist.github.com/vjnadimpalli/083cf423f7b6aa12dcdb14c858bc18a5
Differential Revision: D15772533
Pulled By: vjnadimpalli
fbshipit-source-id: b71dca710590471ede6fb37553388654e2e479b9
Summary:
When using `PRIu64` type of printf specifier, current code base does the following:
```
#ifndef __STDC_FORMAT_MACROS
#define __STDC_FORMAT_MACROS
#endif
#include <inttypes.h>
```
However, this can be simplified to
```
#include <cinttypes>
```
as long as flag `-std=c++11` is used.
This should solve issues like https://github.com/facebook/rocksdb/issues/5159
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5402
Differential Revision: D15701195
Pulled By: miasantreble
fbshipit-source-id: 6dac0a05f52aadb55e9728038599d3d2e4b59d03
Summary:
It's useful to be able to (optionally) associate key-value pairs with user-provided timestamps. This PR is an early effort towards this goal and continues the work of facebook#4942. A suite of new unit tests exist in DBBasicTestWithTimestampWithParam. Support for timestamp requires the user to provide timestamp as a slice in `ReadOptions` and `WriteOptions`. All timestamps of the same database must share the same length, format, etc. The format of the timestamp is the same throughout the same database, and the user is responsible for providing a comparator function (Comparator) to order the <key, timestamp> tuples. Once created, the format and length of the timestamp cannot change (at least for now).
Test plan (on devserver):
```
$COMPILE_WITH_ASAN=1 make -j32 all
$./db_basic_test --gtest_filter=Timestamp/DBBasicTestWithTimestampWithParam.PutAndGet/*
$make check
```
All tests must pass.
We also run the following db_bench tests to verify whether there is regression on Get/Put while timestamp is not enabled.
```
$TEST_TMPDIR=/dev/shm ./db_bench -benchmarks=fillseq,readrandom -num=1000000
$TEST_TMPDIR=/dev/shm ./db_bench -benchmarks=fillrandom -num=1000000
```
Repeat for 6 times for both versions.
Results are as follows:
```
| | readrandom | fillrandom |
| master | 16.77 MB/s | 47.05 MB/s |
| PR5079 | 16.44 MB/s | 47.03 MB/s |
```
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5079
Differential Revision: D15132946
Pulled By: riversand963
fbshipit-source-id: 833a0d657eac21182f0f206c910a6438154c742c
Summary:
util/ means for lower level libraries. trace_replay is highly integrated to DB and sometimes call DB. Move it out to a separate directory.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5376
Differential Revision: D15550938
Pulled By: siying
fbshipit-source-id: f46dce5ceffdc05a73f26379c7bb1b79ebe6c207
Summary:
Many logging related source files are under util/. It will be more structured if they are together.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5387
Differential Revision: D15579036
Pulled By: siying
fbshipit-source-id: 3850134ed50b8c0bb40a0c8ae1f184fa4081303f
Summary:
log_write_bench doesn't compile due to some recent API changes.
This patch fixes the compile by adding the missing params for
OptimizeForLogWrite() and WritableFileWriter().
Signed-off-by: Yuan Zhou <yuan.zhou@intel.com>
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5335
Differential Revision: D15588875
Pulled By: miasantreble
fbshipit-source-id: 726ff4dc227733e915c3b796df25bd3ab0b431ac
Summary:
Right now, with auto roll logger, options.keep_log_file_num enforcement is triggered by events like DB reopen or full obsolete scan happens. In the mean time, the size and number of log files can grow without a limit. We put a stronger enforcement to the option, so that the number of log files can always under control.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5370
Differential Revision: D15570413
Pulled By: siying
fbshipit-source-id: 0916c3c4d42ab8fdd29389ee7fd7e1557b03176e
Summary:
There are too many types of files under util/. Some test related files don't belong to there or just are just loosely related. Mo
ve them to a new directory test_util/, so that util/ is cleaner.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5377
Differential Revision: D15551366
Pulled By: siying
fbshipit-source-id: 0f5c8653832354ef8caa31749c0143815d719e2c
Summary:
util/ means for lower level libraries, so it's a good idea to move the files which requires knowledge to DB out. Create a file/ and move some files there.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5375
Differential Revision: D15550935
Pulled By: siying
fbshipit-source-id: 61a9715dcde5386eebfb43e93f847bba1ae0d3f2
Summary:
The analyzer thinks max_allowed_ space can be 0. In that case, free_space will
be assigned as free_space. It fails to realize that the function call
GetFreeSpace actually sets the free_space variable properly, which is possibly
due to lack of inter-function call analysis.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5365
Differential Revision: D15521790
Pulled By: riversand963
fbshipit-source-id: 839d0a285a1c8773a28a385f0c3be4bb7fbe32cb
Summary:
If RocksDB is configured with a positive max_allowed_space (via sst file manager),
then the sst file manager should use this value instead of total free disk
space to determine whether to clear the background error of space limit
reached.
In DBSSTTest.DBWithMaxSpaceAllowed, we configure a low space limit that is very
likely lower than the free disk space of the test machine. Therefore, once the
test db encounters a Status::SpaceLimit, error handler will call into sst file
manager to start error recovery which may clear the bg error since disk free
space is larger than reserved_disk_buffer_.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5334
Differential Revision: D15501622
Pulled By: riversand963
fbshipit-source-id: 58035efc450b062d6b28c78c322005ec3705fb47
Summary:
Remove PATENTS related wording from a few stragglers which still reference the old PATENTS file.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5326
Differential Revision: D15423297
Pulled By: sagar0
fbshipit-source-id: 4babcddfc120b7d2fed6eb3898287cf8012bf8ea
Summary:
In the current db_bench trace replay, the replay process strictly follows the timestamp to issue the queries. In some cases, user does not care about the time. Therefore, fast forward is needed for users to speed up the replay process.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5273
Differential Revision: D15389232
Pulled By: zhichao-cao
fbshipit-source-id: 735d629b9d2a167b05af3e4fa0ddf9d5d0be1806
Summary:
The recent improvement in https://github.com/facebook/rocksdb/pull/3661 could cause a deadlock: When writing recoverable state, we also commit its sequence number to commit table, which could result into evicting existing commit entry, which could result into advancing max_evicted_seq_, which would need to get snapshots from database, which requires obtaining db mutex. The patch releases db_mutex before calling the callback in WriteRecoverableState to avoid the potential deadlock. It also improves the stress tests to let the issue be manifested in the tests.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5306
Differential Revision: D15341458
Pulled By: maysamyabandeh
fbshipit-source-id: 05dcbed7e21b789fd1e5fd5ee8eea08077162323
Summary:
Part of compaction cpu goes to processing snapshot list, the larger the list the bigger the overhead. Although the lifetime of most of the snapshots is much shorter than the lifetime of compactions, the compaction conservatively operates on the list of snapshots that it initially obtained. This patch allows the snapshot list to be updated via a callback if the compaction is taking long. This should let the compaction to continue more efficiently with much smaller snapshot list.
For simplicity, to avoid the feature is disabled in two cases: i) When more than one sub-compaction are sharing the same snapshot list, ii) when Range Delete is used in which the range delete aggregator has its own copy of snapshot list.
This fixes the reverted https://github.com/facebook/rocksdb/pull/5099 issue with range deletes.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5278
Differential Revision: D15203291
Pulled By: maysamyabandeh
fbshipit-source-id: fa645611e606aa222c7ce53176dc5bb6f259c258
Summary:
Our daily stress tests are failing after this feature. Reverting temporarily until we figure the reason for test failures.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5269
Differential Revision: D15151285
Pulled By: maysamyabandeh
fbshipit-source-id: e4002b99690a97df30d4b4b58bf0f61e9591bc6e
Summary:
Part of compaction cpu goes to processing snapshot list, the larger the list the bigger the overhead. Although the lifetime of most of the snapshots is much shorter than the lifetime of compactions, the compaction conservatively operates on the list of snapshots that it initially obtained. This patch allows the snapshot list to be updated via a callback if the compaction is taking long. This should let the compaction to continue more efficiently with much smaller snapshot list.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5099
Differential Revision: D15086710
Pulled By: maysamyabandeh
fbshipit-source-id: 7649f56c3b6b2fb334962048150142a3bf9c1a12
Summary:
Savepoints are assumed to be used in a stack-wise fashion (only
the top element should be used), so they were stored by `WriteBatch`
in a member variable `save_points` using an std::stack.
Conceptually this is fine, but the implementation had a few issues:
- the `save_points_` instance variable was a plain pointer to a heap-
allocated `SavePoints` struct. The destructor of `WriteBatch` simply
deletes this pointer. However, the copy constructor of WriteBatch
just copied that pointer, meaning that copying a WriteBatch with
active savepoints will very likely have crashed before. Now a proper
copy of the savepoints is made in the copy constructor, and not just
a copy of the pointer
- `save_points_` was an std::stack, which defaults to `std::deque` for
the underlying container. A deque is a bit over the top here, as we
only need access to the most recent savepoint (i.e. stack.top()) but
never any elements at the front. std::deque is rather expensive to
initialize in common environments. For example, the STL implementation
shipped with GNU g++ will perform a heap allocation of more than 500
bytes to create an empty deque object. Although the `save_points_`
container is created lazily by RocksDB, moving from a deque to a plain
`std::vector` is much more memory-efficient. So `save_points_` is now
a vector.
- `save_points_` was changed from a plain pointer to an `std::unique_ptr`,
making ownership more explicit.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5192
Differential Revision: D15024074
Pulled By: maysamyabandeh
fbshipit-source-id: 5b128786d3789cde94e46465c9e91badd07a25d7
Summary:
This PR introduces a new MultiGet() API, with the underlying implementation grouping keys based on SST file and batching lookups in a file. The reason for the new API is twofold - the definition allows callers to allocate storage for status and values on stack instead of std::vector, as well as return values as PinnableSlices in order to avoid copying, and it keeps the original MultiGet() implementation intact while we experiment with batching.
Batching is useful when there is some spatial locality to the keys being queries, as well as larger batch sizes. The main benefits are due to -
1. Fewer function calls, especially to BlockBasedTableReader::MultiGet() and FullFilterBlockReader::KeysMayMatch()
2. Bloom filter cachelines can be prefetched, hiding the cache miss latency
The next step is to optimize the binary searches in the level_storage_info, index blocks and data blocks, since we could reduce the number of key comparisons if the keys are relatively close to each other. The batching optimizations also need to be extended to other formats, such as PlainTable and filter formats. This also needs to be added to db_stress.
Benchmark results from db_bench for various batch size/locality of reference combinations are given below. Locality was simulated by offsetting the keys in a batch by a stride length. Each SST file is about 8.6MB uncompressed and key/value size is 16/100 uncompressed. To focus on the cpu benefit of batching, the runs were single threaded and bound to the same cpu to eliminate interference from other system events. The results show a 10-25% improvement in micros/op from smaller to larger batch sizes (4 - 32).
Batch Sizes
1 | 2 | 4 | 8 | 16 | 32
Random pattern (Stride length 0)
4.158 | 4.109 | 4.026 | 4.05 | 4.1 | 4.074 - Get
4.438 | 4.302 | 4.165 | 4.122 | 4.096 | 4.075 - MultiGet (no batching)
4.461 | 4.256 | 4.277 | 4.11 | 4.182 | 4.14 - MultiGet (w/ batching)
Good locality (Stride length 16)
4.048 | 3.659 | 3.248 | 2.99 | 2.84 | 2.753
4.429 | 3.728 | 3.406 | 3.053 | 2.911 | 2.781
4.452 | 3.45 | 2.833 | 2.451 | 2.233 | 2.135
Good locality (Stride length 256)
4.066 | 3.786 | 3.581 | 3.447 | 3.415 | 3.232
4.406 | 4.005 | 3.644 | 3.49 | 3.381 | 3.268
4.393 | 3.649 | 3.186 | 2.882 | 2.676 | 2.62
Medium locality (Stride length 4096)
4.012 | 3.922 | 3.768 | 3.61 | 3.582 | 3.555
4.364 | 4.057 | 3.791 | 3.65 | 3.57 | 3.465
4.479 | 3.758 | 3.316 | 3.077 | 2.959 | 2.891
dbbench command used (on a DB with 4 levels, 12 million keys)-
TEST_TMPDIR=/dev/shm numactl -C 10 ./db_bench.tmp -use_existing_db=true -benchmarks="readseq,multireadrandom" -write_buffer_size=4194304 -target_file_size_base=4194304 -max_bytes_for_level_base=16777216 -num=12000000 -reads=12000000 -duration=90 -threads=1 -compression_type=none -cache_size=4194304000 -batch_size=32 -disable_auto_compactions=true -bloom_bits=10 -cache_index_and_filter_blocks=true -pin_l0_filter_and_index_blocks_in_cache=true -multiread_batched=true -multiread_stride=4
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5011
Differential Revision: D14348703
Pulled By: anand1976
fbshipit-source-id: 774406dab3776d979c809522a67bedac6c17f84b
Summary:
Introducing Periodic Compactions.
This feature allows all the files in a CF to be periodically compacted. It could help in catching any corruptions that could creep into the DB proactively as every file is constantly getting re-compacted. And also, of course, it helps to cleanup data older than certain threshold.
- Introduced a new option `periodic_compaction_time` to control how long a file can live without being compacted in a CF.
- This works across all levels.
- The files are put in the same level after going through the compaction. (Related files in the same level are picked up as `ExpandInputstoCleanCut` is used).
- Compaction filters, if any, are invoked as usual.
- A new table property, `file_creation_time`, is introduced to implement this feature. This property is set to the time at which the SST file was created (and that time is given by the underlying Env/OS).
This feature can be enabled on its own, or in conjunction with `ttl`. It is possible to set a different time threshold for the bottom level when used in conjunction with ttl. Since `ttl` works only on 0 to last but one levels, you could set `ttl` to, say, 1 day, and `periodic_compaction_time` to, say, 7 days. Since `ttl < periodic_compaction_time` all files in last but one levels keep getting picked up based on ttl, and almost never based on periodic_compaction_time. The files in the bottom level get picked up for compaction based on `periodic_compaction_time`.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5166
Differential Revision: D14884441
Pulled By: sagar0
fbshipit-source-id: 408426cbacb409c06386a98632dcf90bfa1bda47
Summary:
Create new function NPHash64() and GetSliceNPHash64(), which are currently
implemented using murmurhash.
Replace the current direct call of murmurhash() to use the new functions
if the hash results are not used in on-disk format.
This will make it easier to try out or switch to alternative functions
in the uses where data format compatibility doesn't need to be considered.
This part shouldn't have any performance impact.
Also, the sharded cache hash function is changed to the new format, because
it falls into this categoery. It doesn't show visible performance impact
in db_bench results. CPU showed by perf is increased from about 0.2% to 0.4%
in an extreme benchmark setting (4KB blocks, no-compression, everything
cached in block cache). We've known that the current hash function used,
our own Hash() has serious hash quality problem. It can generate a lots of
conflicts with similar input. In this use case, it means extra lock contention
for reads from the same file. This slight CPU regression is worthy to me
to counter the potential bad performance with hot keys. And hopefully this
will get further improved in the future with a better hash function.
cache_test's condition is relaxed a little bit to. The new hash is slightly
more skewed in this use case, but I manually checked the data and see
the hash results are still in a reasonable range.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5155
Differential Revision: D14834821
Pulled By: siying
fbshipit-source-id: ec9a2c0a2f8ae4b54d08b13a5c2e9cc97aa80cb5
Summary:
Annotate all of the logging functions to inform the compiler that these
use printf-style formatting arguments. This allows the compiler to emit
warnings if the format arguments are incorrect.
This also fixes many problems reported now that format string checking
is enabled. Many of these are simply mix-ups in the argument type (e.g,
int vs uint64_t), but in several cases the wrong number of arguments
were being passed in which can cause the code to crash.
The primary motivation for this was to fix the log message in
`DBImpl::SwitchMemtable()` which caused a segfault due to an extra %s
format parameter with no argument supplied.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5089
Differential Revision: D14574795
Pulled By: simpkins
fbshipit-source-id: 0921b03f0743652bf4ae21e414ff54b3bb65422a
Summary:
Since we are planning to use dictionary compression and to use different compression level, it is quite useful to add compression options to TableProperties. For example, in MyRocks, if the feature is available, we can query from information_schema.rocksdb_sst_props to see if all sst files are converted to ZSTD dictionary compressions. Resolves https://github.com/facebook/rocksdb/issues/4992
With this PR, user can query table properties through `GetPropertiesOfAllTables` API and get compression options as std::string:
`window_bits=-14; level=32767; strategy=0; max_dict_bytes=0; zstd_max_train_bytes=0; enabled=0;`
or table_properties->ToString() will also contain it
`# data blocks=1; # entries=13; # deletions=0; # merge operands=0; # range deletions=0; raw key size=143; raw average key size=11.000000; raw value size=39; raw average value size=3.000000; data block size=120; index block size (user-key? 0, delta-value? 0)=27; filter block size=0; (estimated) table size=147; filter policy name=N/A; prefix extractor name=nullptr; column family ID=0; column family name=default; comparator name=leveldb.BytewiseComparator; merge operator name=nullptr; property collectors names=[]; SST file compression algo=Snappy; SST file compression options=window_bits=-14; level=32767; strategy=0; max_dict_bytes=0; zstd_max_train_bytes=0; enabled=0; ; creation time=1552946632; time stamp of earliest key=1552946632;`
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5081
Differential Revision: D14716692
Pulled By: miasantreble
fbshipit-source-id: 7d2f2cf84e052bff876e71b4212cfdebf5be32dd
Summary:
**This PR updates RepeatableThread::wait, breaking some tests on OS X. The rest of the PR fixes the tests on OS X.**
`RepeatableThreadTest.MockEnvTest` uses `MockTimeEnv` and `RepeatableThread`. If `RepeatableThread::wait` calls `TimedWait` with a time smaller than or equal to the current (real) time, `TimedWait` returns immediately on certain platforms, e.g. OS X. #4560 addresses this issue by replacing `TimedWait` with `Wait` in test. This fixes the test but makes test/production code diverge, which is not optimal for test coverage. This PR proposes an alternative fix which unifies test and production code path for `RepeatableThread::wait`. We obtain the current (real) time in seconds and add 10 extra seconds to ensure that `RepeatableThread::wait` invokes `TimedWait` with a time greater than (real) current time. This is to prevent the `TimedWait` function from returning immediately without sleeping and releasing the mutex. If `TimedWait` returns immediately, the mutex will not be released, and `RepeatableThread::TEST_WaitForRun` never has a chance to execute the callback which, in this case, updates the result returned by `mock_env->NowMicros()`. Consequently, `RepeatableThread::wait` cannot break out of the loop, causing test to hang. The extra 10 seconds is a best-effort approach because there seems no reliable and deterministic way to provide the aforementioned guarantee. By the time `RepeatableThread::wait` is called, there is no guarantee that the `delay + mock_env->NowMicros()` will be greater than the current real time. However, 10 seconds should be sufficient in most cases. We will keep an eye for possible flakiness of this test.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5107
Differential Revision: D14680885
Pulled By: riversand963
fbshipit-source-id: d1ecbe10e1dacd110bd464cd01e188bfee72b89e
Summary:
WAL files are currently not subject to deletion rate limiting by DeleteScheduler. If the size of the WAL files is significant, this can cause a high delete rate on SSDs that may affect other operations. To fix it, force WAL file deletions to go through the SstFileManager. Original PR for this is #2768
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5116
Differential Revision: D14669437
Pulled By: anand1976
fbshipit-source-id: c5f62d0640cebaa1574de841a1d01e4ce2faadf0
Summary:
Currently `perf_context.user_key_comparison_count` is bump only in `InternalKeyComparator`. For places user comparator is used directly the counter is not bump. Fixing the majority of it.
Index iterator and filter code also use user comparator directly and don't bump the counter. It is not fixed in this patch.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5098
Differential Revision: D14603753
Pulled By: siying
fbshipit-source-id: 1cd41035644ca9e49b97a51030a5d1e15f5f3cae
Summary:
This PR allows RocksDB to run in single-primary, multi-secondary process mode.
The writer is a regular RocksDB (e.g. an `DBImpl`) instance playing the role of a primary.
Multiple `DBImplSecondary` processes (secondaries) share the same set of SST files, MANIFEST, WAL files with the primary. Secondaries tail the MANIFEST of the primary and apply updates to their own in-memory state of the file system, e.g. `VersionStorageInfo`.
This PR has several components:
1. (Originally in #4745). Add a `PathNotFound` subcode to `IOError` to denote the failure when a secondary tries to open a file which has been deleted by the primary.
2. (Similar to #4602). Add `FragmentBufferedReader` to handle partially-read, trailing record at the end of a log from where future read can continue.
3. (Originally in #4710 and #4820). Add implementation of the secondary, i.e. `DBImplSecondary`.
3.1 Tail the primary's MANIFEST during recovery.
3.2 Tail the primary's MANIFEST during normal processing by calling `ReadAndApply`.
3.3 Tailing WAL will be in a future PR.
4. Add an example in 'examples/multi_processes_example.cc' to demonstrate the usage of secondary RocksDB instance in a multi-process setting. Instructions to run the example can be found at the beginning of the source code.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4899
Differential Revision: D14510945
Pulled By: riversand963
fbshipit-source-id: 4ac1c5693e6012ad23f7b4b42d3c374fecbe8886
Summary:
The stack buffer in rocksdb::autovector is currently defined as an array of elements of the template type. This results in unnecessary construction of those objects, which can be a significant overhead in some cases. This PR changes the type of the stack buf to char* and uses placement new to construct new objects when they are inserted into the autovector.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5080
Differential Revision: D14533221
Pulled By: anand1976
fbshipit-source-id: 9378985c7d03f4e1a28951bdd2403c72f10f23d7
Summary:
Add an option to filter out READ or WRITE operations while tracing.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5082
Differential Revision: D14515083
Pulled By: mrmiywj
fbshipit-source-id: 2504c89a9abf1dd629cad44b4104092702d77610
Summary:
This is a feature to sample data-block compressibility and and report them as stats. 1 in N (tunable) blocks is sampled for compressibility using two algorithms:
1. lz4 or snappy for fast compression
2. zstd or zlib for slow but higher compression.
The stats are reported to the caller as raw-bytes and compressed-bytes. The block continues to be compressed for storage using the specified CompressionType.
The db_bench_tool how has a command line option for specifying the sampling rate. It's default value is 0 (no sampling). To test the overhead for a certain value, users can compare the performance of db_bench_tool, varying the sampling rate. It is unlikely to have a noticeable impact for high values like 20.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4842
Differential Revision: D13629011
Pulled By: shobhitdayal
fbshipit-source-id: 14ca668bcab6499b2a1734edf848eb62a4f4fafa
Summary:
Right now, users can change statistics.stats_level while DB is running, but TSAN may report
data race. We make stats_level_ to be atomic, and access them using accessors.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5030
Differential Revision: D14267519
Pulled By: siying
fbshipit-source-id: 37d7ebeff7a43a406230143422a16af899163f73
Summary:
Statistics cost too much CPU for some use cases. Add two stats levels
so that people can choose to skip two types of expensive stats, timers and
histograms.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5027
Differential Revision: D14252765
Pulled By: siying
fbshipit-source-id: 75ecec9eaa44c06118229df4f80c366115346592
Summary:
When closing a BlobDB, it only waits for background tasks
to finish as the last thing, but the background task may access
some variables that are destroyed. The fix is to introduce a
shutdown function in the timer queue and call the function as
the first thing when destorying BlobDB.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/5005
Differential Revision: D14170342
Pulled By: siying
fbshipit-source-id: 081e6a2d99b9765d5956cf6cdfc290c07270c233
Summary:
This PR adds public `GetStatsHistory` API to retrieve stats history in the form of an std map. The key of the map is the timestamp in microseconds when the stats snapshot is taken, the value is another std map from stats name to stats value (stored in std string). Two DBOptions are introduced: `stats_persist_period_sec` (default 10 minutes) controls the intervals between two snapshots are taken; `max_stats_history_count` (default 10) controls the max number of history snapshots to keep in memory. RocksDB will stop collecting stats snapshots if `stats_persist_period_sec` is set to 0.
(This PR is the in-memory part of https://github.com/facebook/rocksdb/pull/4535)
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4748
Differential Revision: D13961471
Pulled By: miasantreble
fbshipit-source-id: ac836d401ecb84ea92216bf9966f969dedf4ad04
Summary:
The transaction stress tests, stress a high concurrency scenario. In WritePrepared/WriteUnPrepared we need to also stress the scenarios where an inserting/reading transaction is very slow. This would stress the corner cases that the caching is not sufficient and other slower data structures are engaged. To emulate such cases we make use of slow inserter/verifier threads and also reduce the size of cache data structures.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4974
Differential Revision: D14143070
Pulled By: maysamyabandeh
fbshipit-source-id: 81eb674678faf9fae0f654cd60ebcc74e26aeee7
Summary:
MyRocks calls `GetForUpdate` on `INSERT`, for unique key check, and in almost all cases GetForUpdate returns empty result. For such cases, whole key bloom filter is helpful.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4985
Differential Revision: D14118257
Pulled By: miasantreble
fbshipit-source-id: d35cb7109c62fd5ad541a26968e3a3e16d3e85ea
Summary:
We introduced ttl option in CompactionOptionsFIFO when ttl-based file
deletion (compaction) was supported only as part of FIFO Compaction. But
with the extension of ttl semantics even to Level compaction,
CompactionOptionsFIFO.ttl can now be deprecated. Instead we will start
using ColumnFamilyOptions.ttl for FIFO compaction as well.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4965
Differential Revision: D14072960
Pulled By: sagar0
fbshipit-source-id: c98cc2ae695a28136295787cd88d36a220fc219e
Summary:
Make file ingestion atomic.
as title.
Ingesting external SST files into multiple column families should be atomic. If
a crash occurs and db reopens, either all column families have successfully
ingested the files before the crash, or non of the ingestions have any effect
on the state of the db.
Also add unit tests for atomic ingestion.
Note that the unit test here does not cover the case of incomplete atomic group
in the MANIFEST, which is covered in VersionSetTest already.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4895
Differential Revision: D13718245
Pulled By: riversand963
fbshipit-source-id: 7df97cc483af73ad44dd6993008f99b083852198
Summary:
Our previous approach was to train one compression dictionary per compaction, using the first output SST to train a dictionary, and then applying it on subsequent SSTs in the same compaction. While this was great for minimizing CPU/memory/I/O overhead, it did not achieve good compression ratios in practice. In our most promising potential use case, moderate reductions in a dictionary's scope make a major difference on compression ratio.
So, this PR changes compression dictionary to be scoped per-SST. It accepts the tradeoff during table building to use more memory and CPU. Important changes include:
- The `BlockBasedTableBuilder` has a new state when dictionary compression is in-use: `kBuffered`. In that state it accumulates uncompressed data in-memory whenever `Add` is called.
- After accumulating target file size bytes or calling `BlockBasedTableBuilder::Finish`, a `BlockBasedTableBuilder` moves to the `kUnbuffered` state. The transition (`EnterUnbuffered()`) involves sampling the buffered data, training a dictionary, and compressing/writing out all buffered data. In the `kUnbuffered` state, a `BlockBasedTableBuilder` behaves the same as before -- blocks are compressed/written out as soon as they fill up.
- Samples are now whole uncompressed data blocks, except the final sample may be a partial data block so we don't breach the user's configured `max_dict_bytes` or `zstd_max_train_bytes`. The dictionary trainer is supposed to work better when we pass it real units of compression. Previously we were passing 64-byte KV samples which was not realistic.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4952
Differential Revision: D13967980
Pulled By: ajkr
fbshipit-source-id: 82bea6f7537e1529c7a1a4cdee84585f5949300f
Summary:
Enhance ::Insert and ::Verify test functions to add artificial delay between prepare and commit, and take snapshot and reads respectively. A future PR will make use of these to improve stress tests to test against long-running transactions as well as long-running backup jobs. Also randomly sets set_snapshot to false for inserters to skip setting the snapshot in the initialization phase and let the snapshot be taken later explicitly.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4970
Differential Revision: D14031342
Pulled By: maysamyabandeh
fbshipit-source-id: b52b453751f0b25b81b23c48892bc1d152464cab
Summary:
Implement trace sampling to allow user to specify the sampling frequency, i.e. save one per how many requests, so that a user does not need to log all if he/she is interested in only a sampled set.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4963
Differential Revision: D14011190
Pulled By: tang-jianfeng
fbshipit-source-id: 078b631d9319b67cb089dd2c30e21d0df8dc406a
Summary:
Measure CPU time consumed for a compaction and report it in the stats report
Enable NowCPUNanos() to work for MacOS
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4889
Differential Revision: D13701276
Pulled By: zinoale
fbshipit-source-id: 5024e5bbccd4dd10fd90d947870237f436445055
Summary:
Blob DB files are not tracked by the SFM, so they currently don't get
deleted in the background. Force them to be deleted in background so
rate limiting can be applied
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4928
Differential Revision: D13854649
Pulled By: anand1976
fbshipit-source-id: 8031ce66842ff0af440c715d886b377983dad7d8
Summary:
I didn't find where customized hash function is used in DynamicBloom. This can only reduce performance. Remove it.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4915
Differential Revision: D13794452
Pulled By: siying
fbshipit-source-id: e38669b11e01444d2d782da11c7decabbd851819
Summary:
- If block cache disabled or not used for meta-blocks, `BlockBasedTableReader::Rep::uncompression_dict` owns the `UncompressionDict`. It is preloaded during `PrefetchIndexAndFilterBlocks`.
- If block cache is enabled and used for meta-blocks, block cache owns the `UncompressionDict`, which holds dictionary and digested dictionary when needed. It is never prefetched though there is a TODO for this in the code. The cache key is simply the compression dictionary block handle.
- New stats for compression dictionary accesses in block cache: "BLOCK_CACHE_COMPRESSION_DICT_*" and "compression_dict_block_read_count"
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4881
Differential Revision: D13663801
Pulled By: ajkr
fbshipit-source-id: bdcc54044e180855cdcc57639b493b0e016c9a3f
Summary:
Right now, deleting blob files is not rate limited, even if SstFileManger is specified.
On the other hand, rate limiting blob deletion is not supported. With this change, Blob file
deletion will go through SstFileManager too.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4904
Differential Revision: D13772545
Pulled By: siying
fbshipit-source-id: bd1b1d0beb26d5167385e00b7ecb8b94b879de84
Summary:
This is essentially a re-submission of #4251 with a few improvements:
- Split `CompressionDict` into two separate classes: `CompressionDict` and `UncompressionDict`
- Eliminated `Init` functions. Instead do all initialization work in constructors.
- Added test case for parallel DB open, which is the scenario where #4251 failed under TSAN.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4849
Differential Revision: D13606039
Pulled By: ajkr
fbshipit-source-id: 08c236059798c710db9cbf545fce0f371232d447
Summary:
By convention, time_t almost always stores the integral number of seconds since
00:00 hours, Jan 1, 1970 UTC, according to http://www.cplusplus.com/reference/ctime/time_t/.
We surely want more precision than seconds.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4868
Differential Revision: D13633046
Pulled By: riversand963
fbshipit-source-id: 4e01e23a22e8838023c51a91247a286dbf3a5396
Summary:
Declare Jemalloc non-standard APIs as weak symbols, so that if Jemalloc is linked with the binary, these symbols will be replaced by Jemalloc's, otherwise they will be nullptr. This is similar to how folly detect jemalloc, but we assume the main program use jemalloc as long as jemalloc is linked: https://github.com/facebook/folly/blob/master/folly/memory/Malloc.h#L147
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4844
Differential Revision: D13574934
Pulled By: yiwu-arbug
fbshipit-source-id: 7ea871beb1be7d5a1259cc38f9b78078793db2db
Summary:
1. Remove unused API SubtractCompactionTask().
2. Assert outstanding tasks drop to zero in ConcurrentTaskLimiterImpl destructor.
3. Remove GetOutstandingTask() check from manual compaction test, as TEST_WaitForCompact() doesn't synced with 'delete prepicked_compaction' in DBImpl::BGWorkCompaction(), which may make the test flaky.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4795
Differential Revision: D13542183
Pulled By: siying
fbshipit-source-id: 5eb2a47e62efe4126937149aa0df6e243ebefc33
Summary:
RangeDelAggregatorV2 now supports ShouldDelete calls on
snapshot stripes and creation of range tombstone compaction iterators.
RangeDelAggregator is no longer used on any non-test code path, and will
be removed in a future commit.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4758
Differential Revision: D13439254
Pulled By: abhimadan
fbshipit-source-id: fe105bcf8e3d4a2df37a622d5510843cd71b0401
Summary:
The PR is targeting to resolve the issue of:
https://github.com/facebook/rocksdb/issues/3972#issue-330771918
We have a rocksdb created with leveled-compaction with multiple column families (CFs), some of CFs are using HDD to store big and less frequently accessed data and others are using SSD.
When there are continuously write traffics going on to all CFs, the compaction thread pool is mostly occupied by those slow HDD compactions, which blocks fully utilize SSD bandwidth.
Since atomic write and transaction is needed across CFs, so splitting it to multiple rocksdb instance is not an option for us.
With the compaction thread control, we got 30%+ HDD write throughput gain, and also a lot smooth SSD write since less write stall happening.
ConcurrentTaskLimiter can be shared with multi-CFs across rocksdb instances, so the feature does not only work for multi-CFs scenarios, but also for multi-rocksdbs scenarios, who need disk IO resource control per tenant.
The usage is straight forward:
e.g.:
//
// Enable compaction thread limiter thru ColumnFamilyOptions
//
std::shared_ptr<ConcurrentTaskLimiter> ctl(NewConcurrentTaskLimiter("foo_limiter", 4));
Options options;
ColumnFamilyOptions cf_opt(options);
cf_opt.compaction_thread_limiter = ctl;
...
//
// Compaction thread limiter can be tuned or disabled on-the-fly
//
ctl->SetMaxOutstandingTask(12); // enlarge to 12 tasks
...
ctl->ResetMaxOutstandingTask(); // disable (bypass) thread limiter
ctl->SetMaxOutstandingTask(-1); // Same as above
...
ctl->SetMaxOutstandingTask(0); // full throttle (0 task)
//
// Sharing compaction thread limiter among CFs (to resolve multiple storage perf issue)
//
std::shared_ptr<ConcurrentTaskLimiter> ctl_ssd(NewConcurrentTaskLimiter("ssd_limiter", 8));
std::shared_ptr<ConcurrentTaskLimiter> ctl_hdd(NewConcurrentTaskLimiter("hdd_limiter", 4));
Options options;
ColumnFamilyOptions cf_opt_ssd1(options);
ColumnFamilyOptions cf_opt_ssd2(options);
ColumnFamilyOptions cf_opt_hdd1(options);
ColumnFamilyOptions cf_opt_hdd2(options);
ColumnFamilyOptions cf_opt_hdd3(options);
// SSD CFs
cf_opt_ssd1.compaction_thread_limiter = ctl_ssd;
cf_opt_ssd2.compaction_thread_limiter = ctl_ssd;
// HDD CFs
cf_opt_hdd1.compaction_thread_limiter = ctl_hdd;
cf_opt_hdd2.compaction_thread_limiter = ctl_hdd;
cf_opt_hdd3.compaction_thread_limiter = ctl_hdd;
...
//
// The limiter is disabled by default (or set to nullptr explicitly)
//
Options options;
ColumnFamilyOptions cf_opt(options);
cf_opt.compaction_thread_limiter = nullptr;
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4332
Differential Revision: D13226590
Pulled By: siying
fbshipit-source-id: 14307aec55b8bd59c8223d04aa6db3c03d1b0c1d
Summary:
In Direct I/O case, WritableFileWriter::Close() rewrites the last block again, even if there is nothing new. The reason is that, Close() flushes the buffer. For non-direct I/O case, the buffer is empty in this case so it is a no-op. However, in direct I/O case, the partial data in the last block is kept in the buffer because it needs to be rewritten for the next write. This piece of data is flushed again. This commit fixes it by skipping this write out if `pending_sync_` flag shows that there isn't new data sync last sync.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4771
Differential Revision: D13420426
Pulled By: siying
fbshipit-source-id: 9d39ec9a215b1425d4ed40d85e0eba1f5daa75c6
Summary:
Transaction::GetForUpdate is extended with a do_validate parameter with default value of true. If false it skips validating the snapshot (if there is any) before doing the read. After the read it also returns the latest value (expects the ReadOptions::snapshot to be nullptr). This allows RocksDB applications to use GetForUpdate similarly to how InnoDB does. Similarly ::Merge, ::Put, ::Delete, and ::SingleDelete are extended with assume_exclusive_tracked with default value of false. It true it indicates that call is assumed to be after a ::GetForUpdate(do_validate=false).
The Java APIs are accordingly updated.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4680
Differential Revision: D13068508
Pulled By: maysamyabandeh
fbshipit-source-id: f0b59db28f7f6a078b60844d902057140765e67d
Summary:
Currently tests are failing on master with the following message:
> util/jemalloc_nodump_allocator.cc:132:8: error: unused parameter ‘options’ [-Werror=unused-parameter]
Status NewJemallocNodumpAllocator(
This PR attempts to fix the issue
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4738
Differential Revision: D13278804
Pulled By: miasantreble
fbshipit-source-id: 64a6204aa685bd85d8b5080655cafef9980fac2f
Summary:
Add option to limit tcache usage by allocation size. This is to reduce total tcache size in case there are many user threads accessing the allocator and incur non-trivial memory usage.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4736
Differential Revision: D13269305
Pulled By: yiwu-arbug
fbshipit-source-id: 95a9b7fc67facd66837c849137e30e137112e19d
Summary:
Removed `one_time_use` flag, which removed the need for some
tests, and changed all `NewRangeTombstoneIterator` methods to return
`FragmentedRangeTombstoneIterators`.
These changes also led to removing `RangeDelAggregatorV2::AddUnfragmentedTombstones`
and one of the `MemTableListVersion::AddRangeTombstoneIterators` methods.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4692
Differential Revision: D13106570
Pulled By: abhimadan
fbshipit-source-id: cbab5432d7fc2d9cdfd8d9d40361a1bffaa8f845
Summary:
If user do not end the trace manually, the tracing will continue which can potential use up all the storage space and cause problem. In this PR, the max trace file size is added to the TraceOptions and user can set the value if they need or the default is 64GB.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4610
Differential Revision: D12893400
Pulled By: zhichao-cao
fbshipit-source-id: acf4b5a6076bb691778bdfbac4864e1006758953
Summary:
Previously, every range tombstone iterator was seeked on every
ShouldDelete call, which quickly degraded performance for long range
scans. This PR improves performance by tracking iterator positions and
only advancing iterators when necessary.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4677
Differential Revision: D13205373
Pulled By: abhimadan
fbshipit-source-id: 80c199dace1e19362a4c61c686bf01913eae87cb
Summary:
It called the autovector::push_back simply in autovector::emplace_back.
This was not efficient, and then optimazed this function through the
perfect forwarding.
This was the src and result of the benchmark(using the google'benchmark library, the type of elem in
autovector was std::string, and call emplace_back with the "char *" type):
https://gist.github.com/monadbobo/93448b89a42737b08cbada81de75c5cd
PS: The benchmark's result of previous PR was not accurate, and so I update the test case and result.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4606
Differential Revision: D13046813
Pulled By: sagar0
fbshipit-source-id: 19cde1bcadafe899aa454b703acb35737a1cc02d
Summary:
Use the `DBOptions` that the backup engine already holds to figure out the right `EnvOptions` to use when reading the DB files. This means that, if a user opened a DB instance with `use_direct_reads=true`, then using `BackupEngine` to back up that DB instance will use direct I/O to read files when calculating checksums and copying. Currently the WALs and manifests would still be read using buffered I/O to prevent mixing direct I/O reads with concurrent buffered I/O writes.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4640
Differential Revision: D13015268
Pulled By: ajkr
fbshipit-source-id: 77006ad6f3e00ce58374ca4793b785eea0db6269
Summary:
Ran the following commands to recursively change all the files under RocksDB:
```
find . -type f -name "*.cc" -exec sed -i 's/ unique_ptr/ std::unique_ptr/g' {} +
find . -type f -name "*.cc" -exec sed -i 's/<unique_ptr/<std::unique_ptr/g' {} +
find . -type f -name "*.cc" -exec sed -i 's/ shared_ptr/ std::shared_ptr/g' {} +
find . -type f -name "*.cc" -exec sed -i 's/<shared_ptr/<std::shared_ptr/g' {} +
```
Running `make format` updated some formatting on the files touched.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4638
Differential Revision: D12934992
Pulled By: sagar0
fbshipit-source-id: 45a15d23c230cdd64c08f9c0243e5183934338a8
Summary:
Long absolute file names in log make it hard to read the LOG files.
So we shorter them to relative to the root of RocksDB project path.
In most cases, they will only have one level directory and one file name.
There was [a talk](#4316) about making "util/logging.h" a public header file.
But we concern the conflicts that might be introduced in for macros
named `STRINGIFY`, `TOSTRING`, and `PREPEND_FILE_LINE`.
So I prepend a prefix `ROCKS_LOG_` to them.
I also remove the line that includes "port.h" which seems unneccessary here.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4616
Differential Revision: D12892857
Pulled By: siying
fbshipit-source-id: af79aaf82153b8fd66b5966aced39a51fbca9c6c
Summary:
Introduce `JemallocNodumpAllocator`, which allow exclusion of block cache usage from core dump. It utilize custom hook of jemalloc arena, and when jemalloc arena request memory from system, the allocator use the hook to set `MADV_DONTDUMP ` to the memory. The implementation is basically the same as `folly::JemallocNodumpAllocator`, except for some minor difference:
1. It only support jemalloc >= 5.0
2. When the allocator destruct, it explicitly destruct the corresponding arena via `arena.<i>.destroy` via `mallctl`.
Depending on #4502.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4534
Differential Revision: D10435474
Pulled By: yiwu-arbug
fbshipit-source-id: e80edea755d3853182485d2be710376384ce0bb4
Summary:
Rename the interface, as it is mean to be a generic interface for memory allocation.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4590
Differential Revision: D10866340
Pulled By: yiwu-arbug
fbshipit-source-id: 85cb753351a40cb856c046aeaa3f3b369eef3d16
Summary:
Previously, range tombstones were accumulated from every level, which
was necessary if a range tombstone in a higher level covered a key in a lower
level. However, RangeDelAggregator::AddTombstones's complexity is based on
the number of tombstones that are currently stored in it, which is wasteful in
the Get case, where we only need to know the highest sequence number of range
tombstones that cover the key from higher levels, and compute the highest covering
sequence number at the current level. This change introduces this optimization, and
removes the use of RangeDelAggregator from the Get path.
In the benchmark results, the following command was used to initialize the database:
```
./db_bench -db=/dev/shm/5k-rts -use_existing_db=false -benchmarks=filluniquerandom -write_buffer_size=1048576 -compression_type=lz4 -target_file_size_base=1048576 -max_bytes_for_level_base=4194304 -value_size=112 -key_size=16 -block_size=4096 -level_compaction_dynamic_level_bytes=true -num=5000000 -max_background_jobs=12 -benchmark_write_rate_limit=20971520 -range_tombstone_width=100 -writes_per_range_tombstone=100 -max_num_range_tombstones=50000 -bloom_bits=8
```
...and the following command was used to measure read throughput:
```
./db_bench -db=/dev/shm/5k-rts/ -use_existing_db=true -benchmarks=readrandom -disable_auto_compactions=true -num=5000000 -reads=100000 -threads=32
```
The filluniquerandom command was only run once, and the resulting database was used
to measure read performance before and after the PR. Both binaries were compiled with
`DEBUG_LEVEL=0`.
Readrandom results before PR:
```
readrandom : 4.544 micros/op 220090 ops/sec; 16.9 MB/s (63103 of 100000 found)
```
Readrandom results after PR:
```
readrandom : 11.147 micros/op 89707 ops/sec; 6.9 MB/s (63103 of 100000 found)
```
So it's actually slower right now, but this PR paves the way for future optimizations (see #4493).
----
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4449
Differential Revision: D10370575
Pulled By: abhimadan
fbshipit-source-id: 9a2e152be1ef36969055c0e9eb4beb0d96c11f4d
Summary:
This fixes three tests that fail with relatively recent tools and libraries:
The tests are:
* `spatial_db_test`
* `table_test`
* `db_universal_compaction_test`
I'm using:
* `gcc` 7.3.0
* `glibc` 2.27
* `snappy` 1.1.7
* `gflags` 2.2.1
* `zlib` 1.2.11
* `bzip2` 1.0.6.0.1
* `lz4` 1.8.2
* `jemalloc` 5.0.1
The versions used in the Travis environment (which is two Ubuntu LTS versions behind the current one and doesn't use `lz4` or `jemalloc`) don't seem to have a problem. However, to be safe, I verified that these tests pass with and without my changes in a trusty Docker container without `lz4` and `jemalloc`.
However, I do get an unrelated set of other failures when using a trusty Docker container that uses `lz4` and `jemalloc`:
```
db/db_universal_compaction_test.cc:506: Failure
Value of: num + 1
Actual: 3
Expected: NumSortedRuns(1)
Which is: 4
[ FAILED ] UniversalCompactionNumLevels/DBTestUniversalCompaction.DynamicUniversalCompactionReadAmplification/0, where GetParam() = (1, false) (1189 ms)
[ RUN ] UniversalCompactionNumLevels/DBTestUniversalCompaction.DynamicUniversalCompactionReadAmplification/1
db/db_universal_compaction_test.cc:506: Failure
Value of: num + 1
Actual: 3
Expected: NumSortedRuns(1)
Which is: 4
[ FAILED ] UniversalCompactionNumLevels/DBTestUniversalCompaction.DynamicUniversalCompactionReadAmplification/1, where GetParam() = (1, true) (1246 ms)
[ RUN ] UniversalCompactionNumLevels/DBTestUniversalCompaction.DynamicUniversalCompactionReadAmplification/2
db/db_universal_compaction_test.cc:506: Failure
Value of: num + 1
Actual: 3
Expected: NumSortedRuns(1)
Which is: 4
[ FAILED ] UniversalCompactionNumLevels/DBTestUniversalCompaction.DynamicUniversalCompactionReadAmplification/2, where GetParam() = (3, false) (1237 ms)
[ RUN ] UniversalCompactionNumLevels/DBTestUniversalCompaction.DynamicUniversalCompactionReadAmplification/3
db/db_universal_compaction_test.cc:506: Failure
Value of: num + 1
Actual: 3
Expected: NumSortedRuns(1)
Which is: 4
[ FAILED ] UniversalCompactionNumLevels/DBTestUniversalCompaction.DynamicUniversalCompactionReadAmplification/3, where GetParam() = (3, true) (1195 ms)
[ RUN ] UniversalCompactionNumLevels/DBTestUniversalCompaction.DynamicUniversalCompactionReadAmplification/4
db/db_universal_compaction_test.cc:506: Failure
Value of: num + 1
Actual: 3
Expected: NumSortedRuns(1)
Which is: 4
[ FAILED ] UniversalCompactionNumLevels/DBTestUniversalCompaction.DynamicUniversalCompactionReadAmplification/4, where GetParam() = (5, false) (1161 ms)
[ RUN ] UniversalCompactionNumLevels/DBTestUniversalCompaction.DynamicUniversalCompactionReadAmplification/5
db/db_universal_compaction_test.cc:506: Failure
Value of: num + 1
Actual: 3
Expected: NumSortedRuns(1)
Which is: 4
[ FAILED ] UniversalCompactionNumLevels/DBTestUniversalCompaction.DynamicUniversalCompactionReadAmplification/5, where GetParam() = (5, true) (1229 ms)
```
I haven't attempted to fix these since I'm not using trusty and Travis doesn't use `lz4` and `jemalloc`. However, the final commit in this PR does at least fix the compilation errors that occur when using trusty's version of `lz4`.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4562
Differential Revision: D10510917
Pulled By: maysamyabandeh
fbshipit-source-id: 59534042015ec339270e5fc2f6ac4d859370d189
Summary:
A fix similar to #4410 but on the write path. On IO error on `SelectBlobFile()` we didn't return error code properly, but simply a nullptr of `BlobFile`. The `AppendBlob()` method didn't have null check for the pointer and caused crash. The fix make sure we properly return error code in this case.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4580
Differential Revision: D10513849
Pulled By: yiwu-arbug
fbshipit-source-id: 80bca920d1d7a3541149de981015ad83e0aa14b5
Summary:
Couple of very minor improvements (typos in comments, full qualification of class name, reordering members of a struct to make it smaller)
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4564
Differential Revision: D10510183
Pulled By: maysamyabandeh
fbshipit-source-id: c7ddf9bfbf2db08cd31896c3fd93789d3fa68c8b
Summary:
When `MockTimeEnv` is used in test to mock time methods, we cannot use `CondVar::TimedWait` because it is using real time, not the mocked time for wait timeout. On Mac the method can return immediately without awaking other waiting threads, if the real time is larger than `wait_until` (which is a mocked time). When that happen, the `wait()` method will fall into an infinite loop.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4560
Differential Revision: D10472851
Pulled By: yiwu-arbug
fbshipit-source-id: 898902546ace7db7ac509337dd8677a527209d19
Summary:
Current `log::Reader` does not perform retry after encountering `EOF`. In the future, we need the log reader to be able to retry tailing the log even after `EOF`.
Current implementation is simple. It does not provide more advanced retry policies. Will address this in the future.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4394
Differential Revision: D9926508
Pulled By: riversand963
fbshipit-source-id: d86d145792a41bd64a72f642a2a08c7b7b5201e1
Summary:
Using const string& can avoid one extra string copy. This PR addresses a recent comment made by siying on #3933.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4491
Differential Revision: D10381211
Pulled By: riversand963
fbshipit-source-id: 27fc2d65d84bc7cd07833c77cdc47f06dcfaeb31
Summary:
We would like to collect file-system-level statistics including file name, offset, length, return code, latency, etc., which requires to add callbacks to intercept file IO function calls when RocksDB is running.
To collect file-system-level statistics, users can inherit the class `EventListener`, as in `TestFileOperationListener `. Note that `TestFileOperationListener::ShouldBeNotifiedOnFileIO()` returns true.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/3933
Differential Revision: D10219571
Pulled By: riversand963
fbshipit-source-id: 7acc577a2d31097766a27adb6f78eaf8b1e8ff15
Summary:
Closes https://github.com/facebook/rocksdb/issues/4462
I'm not sure if you'll be happy with `std::random_device{}`, perhaps you would want to use your rand instance instead. I didn't test to see if your rand instance supports the requirements that `std::shuffle` takes.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4482
Differential Revision: D10325133
Pulled By: yiwu-arbug
fbshipit-source-id: 47b7adaf4bb2b8d64cf090ea6b1b48ef53180581
Summary:
The original logic was assuming that the only architectures that the code would build for on Windows were x86 and x64. This change will enable building for arm32 on Windows as well.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4349
Differential Revision: D10280887
Pulled By: sagar0
fbshipit-source-id: 9ca0bede25505d22e13acf916d38aeeaaf5d981a
Summary:
It also renames InstallMemtableFlushResults to MaybeInstallMemtableFlushResults to clarify its contract.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4464
Differential Revision: D10224918
Pulled By: maysamyabandeh
fbshipit-source-id: 04e3f2d8542002cb9f8010cb436f5152751b3cbe
Summary:
The old error message was misleading because it led people to believe the truncation operation failed.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4454
Differential Revision: D10203575
Pulled By: riversand963
fbshipit-source-id: c76482a132566635cb55d4c73d45c461f295ec43
Summary:
This is a conceptually simple change, but it touches many files to
pass the allocator through function calls.
We introduce CacheAllocator, which can be used by clients to configure
custom allocator for cache blocks. Our motivation is to hook this up
with folly's `JemallocNodumpAllocator`
(f43ce6d686/folly/experimental/JemallocNodumpAllocator.h),
but there are many other possible use cases.
Additionally, this commit cleans up memory allocation in
`util/compression.h`, making sure that all allocations are wrapped in a
unique_ptr as soon as possible.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4437
Differential Revision: D10132814
Pulled By: yiwu-arbug
fbshipit-source-id: be1343a4b69f6048df127939fea9bbc96969f564
Summary:
Introduce `RepeatableThread` utility to run task periodically in a separate thread. It is basically the same as the the same class in fbcode, and in addition provide a helper method to let tests mock time and trigger execution one at a time.
We can use this class to replace `TimerQueue` in #4382 and `BlobDB`.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4423
Differential Revision: D10020932
Pulled By: yiwu-arbug
fbshipit-source-id: 3616bef108c39a33c92eedb1256de424b7c04087
Summary:
Fix IO error on read not being handle and crashing the DB. With the fix we properly return the error.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4410
Differential Revision: D9979246
Pulled By: yiwu-arbug
fbshipit-source-id: 111a85675067a29c03cb60e9a34103f4ff636694
Summary:
1. Add override keyword to overridden virtual functions in EventListener
2. Fix a memory corruption that can happen during DB shutdown when in
read-only mode due to a background write error
3. Fix uninitialized buffers in error_handler_test.cc that cause
valgrind to complain
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4375
Differential Revision: D9875779
Pulled By: anand1976
fbshipit-source-id: 022ede1edc01a9f7e21ecf4c61ef7d46545d0640
Summary:
This commit implements automatic recovery from a Status::NoSpace() error
during background operations such as write callback, flush and
compaction. The broad design is as follows -
1. Compaction errors are treated as soft errors and don't put the
database in read-only mode. A compaction is delayed until enough free
disk space is available to accomodate the compaction outputs, which is
estimated based on the input size. This means that users can continue to
write, and we rely on the WriteController to delay or stop writes if the
compaction debt becomes too high due to persistent low disk space
condition
2. Errors during write callback and flush are treated as hard errors,
i.e the database is put in read-only mode and goes back to read-write
only fater certain recovery actions are taken.
3. Both types of recovery rely on the SstFileManagerImpl to poll for
sufficient disk space. We assume that there is a 1-1 mapping between an
SFM and the underlying OS storage container. For cases where multiple
DBs are hosted on a single storage container, the user is expected to
allocate a single SFM instance and use the same one for all the DBs. If
no SFM is specified by the user, DBImpl::Open() will allocate one, but
this will be one per DB and each DB will recover independently. The
recovery implemented by SFM is as follows -
a) On the first occurance of an out of space error during compaction,
subsequent
compactions will be delayed until the disk free space check indicates
enough available space. The required space is computed as the sum of
input sizes.
b) The free space check requirement will be removed once the amount of
free space is greater than the size reserved by in progress
compactions when the first error occured
c) If the out of space error is a hard error, a background thread in
SFM will poll for sufficient headroom before triggering the recovery
of the database and putting it in write-only mode. The headroom is
calculated as the sum of the write_buffer_size of all the DB instances
associated with the SFM
4. EventListener callbacks will be called at the start and completion of
automatic recovery. Users can disable the auto recov ery in the start
callback, and later initiate it manually by calling DB::Resume()
Todo:
1. More extensive testing
2. Add disk full condition to db_stress (follow-on PR)
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4164
Differential Revision: D9846378
Pulled By: anand1976
fbshipit-source-id: 80ea875dbd7f00205e19c82215ff6e37da10da4a
Summary:
Reverting is needed to unblock a user building against master, who is blocked for multiple days due to a thread-safety issue in `GetEmptyDict`. We haven't been able to fix it quickly, so reverting.
Simply ran `git revert 6c40806e51a89386d2b066fddf73d3fd03a36f65`. There were no merge conflicts.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4347
Differential Revision: D9668365
Pulled By: ajkr
fbshipit-source-id: 0c56334f0a23cf5ee0233d4e4679eae6709739cd
Summary:
As you know, almost all compilers support "pragma once" keyword instead of using include guards. To be keep consistency between header files, all header files are edited.
Besides this, try to fix some warnings about loss of data.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4339
Differential Revision: D9654990
Pulled By: ajkr
fbshipit-source-id: c2cf3d2d03a599847684bed81378c401920ca848
Summary:
`GetLiveFiles` and `GetLiveFilesMetadata` should return path relative to db path.
It is a separate issue when `path_relative` is false how can we return relative path. But `DBImpl::GetLiveFiles` don't handle it as well when there are multiple `db_paths`.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4326
Differential Revision: D9545904
Pulled By: yiwu-arbug
fbshipit-source-id: 6762d879fcb561df2b612e6fdfb4a6b51db03f5d
Summary: For the CURRENT file forged during checkpoint, we were forgetting to `fsync` or `fdatasync` it after its creation. This PR fixes it.
Differential Revision: D9525939
Pulled By: ajkr
fbshipit-source-id: a505483644026ee3f501cfc0dcbe74832165b2e3
Summary:
In RocksDB, for a given SST file, all data blocks are compressed with the same dictionary. When we compress a block using the dictionary's raw bytes, the compression library first has to digest the dictionary to get it into a usable form. This digestion work is redundant and ideally should be done once per file.
ZSTD offers APIs for the caller to create and reuse a digested dictionary object (`ZSTD_CDict`). In this PR, we call `ZSTD_createCDict` once per file to digest the raw bytes. Then we use `ZSTD_compress_usingCDict` to compress each data block using the pre-digested dictionary. Once the file's created `ZSTD_freeCDict` releases the resources held by the digested dictionary.
There are a couple other changes included in this PR:
- Changed the parameter object for (un)compression functions from `CompressionContext`/`UncompressionContext` to `CompressionInfo`/`UncompressionInfo`. This avoids the previous pattern, where `CompressionContext`/`UncompressionContext` had to be mutated before calling a (un)compression function depending on whether dictionary should be used. I felt that mutation was error-prone so eliminated it.
- Added support for digested uncompression dictionaries (`ZSTD_DDict`) as well. However, this PR does not support reusing them across uncompression calls for the same file. That work is deferred to a later PR when we will store the `ZSTD_DDict` objects in block cache.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4251
Differential Revision: D9257078
Pulled By: ajkr
fbshipit-source-id: 21b8cb6bbdd48e459f1c62343780ab66c0a64438
Summary:
We want to sample the file I/O issued by RocksDB and report the function calls. This requires us to include the file paths otherwise it's hard to tell what has been going on.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4039
Differential Revision: D8670178
Pulled By: riversand963
fbshipit-source-id: 97ee806d1c583a2983e28e213ee764dc6ac28f7a
Summary:
ZSTD's dynamic library exports `ZDICT_trainFromBuffer` symbol since v1.1.3, and its static library exports it since v0.6.1. We don't know whether linkage is static or dynamic, so just require v1.1.3 to use dictionary trainer.
Fixes the issue reported here: https://jira.mariadb.org/browse/MDEV-16525.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4295
Differential Revision: D9417183
Pulled By: ajkr
fbshipit-source-id: 0e89d2f48d9e7f6eee73e7f4572660a9f7122db8
Summary:
Clang analyze is not happy in two pieces of code, with "Potential memory leak". No idea what the problem but slightly changing the code makes clang happy.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4292
Differential Revision: D9413555
Pulled By: siying
fbshipit-source-id: 9428c9d3664530c72129feefd135ee63d8386137
Summary:
Add hash index support to data blocks, which helps to reduce the CPU utilization of point-lookup operations. This feature is backward compatible with the data block created without the hash index. It is disabled by default unless `BlockBasedTableOptions::data_block_index_type` is set to `data_block_index_type = kDataBlockBinaryAndHash.`
The DB size would be bigger with the hash index option as a hash table is added at the end of each data block. If the hash utilization ratio is 1:1, the space overhead is one byte per key. The hash table utilization ratio is adjustable using `BlockBasedTableOptions::data_block_hash_table_util_ratio`. A lower utilization ratio will improve more on the point-lookup efficiency, but take more space too.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4174
Differential Revision: D8965914
Pulled By: fgwu
fbshipit-source-id: 1c6bae5d1fc39c80282d8890a72e9e67bc247198
Summary:
A framework of trace analyzing for RocksDB
After collecting the trace by using the tool of [PR #3837](https://github.com/facebook/rocksdb/pull/3837). User can use the Trace Analyzer to interpret, analyze, and characterize the collected workload.
**Input:**
1. trace file
2. Whole keys space file
**Statistics:**
1. Access count of each operation (Get, Put, Delete, SingleDelete, DeleteRange, Merge) in each column family.
2. Key hotness (access count) of each one
3. Key space separation based on given prefix
4. Key size distribution
5. Value size distribution if appliable
6. Top K accessed keys
7. QPS statistics including the average QPS and peak QPS
8. Top K accessed prefix
9. The query correlation analyzing, output the number of X after Y and the corresponding average time
intervals
**Output:**
1. key access heat map (either in the accessed key space or whole key space)
2. trace sequence file (interpret the raw trace file to line base text file for future use)
3. Time serial (The key space ID and its access time)
4. Key access count distritbution
5. Key size distribution
6. Value size distribution (in each intervals)
7. whole key space separation by the prefix
8. Accessed key space separation by the prefix
9. QPS of each operation and each column family
10. Top K QPS and their accessed prefix range
**Test:**
1. Added the unit test of analyzing Get, Put, Delete, SingleDelete, DeleteRange, Merge
2. Generated the trace and analyze the trace
**Implemented but not tested (due to the limitation of trace_replay):**
1. Analyzing Iterator, supporting Seek() and SeekForPrev() analyzing
2. Analyzing the number of Key found by Get
**Future Work:**
1. Support execution time analyzing of each requests
2. Support cache hit situation and block read situation of Get
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4091
Differential Revision: D9256157
Pulled By: zhichao-cao
fbshipit-source-id: f0ceacb7eedbc43a3eee6e85b76087d7832a8fe6
Summary:
In the current trace_and replay, Get an WriteBatch are traced. This pull request track down the Seek() and SeekForPrev() to the trace file. <target_key, timestamp, column_family_id> are write to the file.
Replay of Iterator is not supported in the current implementation.
Tested with trace_analyzer.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4228
Differential Revision: D9201381
Pulled By: zhichao-cao
fbshipit-source-id: 6f9cc9cb3c20260af741bee065ec35c5c96354ab
Summary:
Given that index value is a BlockHandle, which is basically an <offset, size> pair we can apply delta encoding on the values. The first value at each index restart interval encoded the full BlockHandle but the rest encode only the size. Refer to IndexBlockIter::DecodeCurrentValue for the detail of the encoding. This reduces the index size which helps using the block cache more efficiently. The feature is enabled with using format_version 4.
The feature comes with a bit of cpu overhead which should be paid back by the higher cache hits due to smaller index block size.
Results with sysbench read-only using 4k blocks and using 16 index restart interval:
Format 2:
19585 rocksdb read-only range=100
Format 3:
19569 rocksdb read-only range=100
Format 4:
19352 rocksdb read-only range=100
Pull Request resolved: https://github.com/facebook/rocksdb/pull/3983
Differential Revision: D8361343
Pulled By: maysamyabandeh
fbshipit-source-id: f882ee082322acac32b0072e2bdbb0b5f854e651
Summary:
HashMayMatch is related to AddKey() instead of CreateFilter().
Also applies some minor Fixes#4191#4200#3910
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4202
Differential Revision: D9180945
Pulled By: maysamyabandeh
fbshipit-source-id: 6f07b81c5bb9bda5c0273475b486ba8a030471e6
Summary:
Hi, it would be great if we could expose this API, so that LogDevice can use it to track the total size of trash files and alarm if it grows too large in relation to disk size. There's probably other customers that would be interested in this as well. :)
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4206
Differential Revision: D9115516
Pulled By: gdavidsson
fbshipit-source-id: f34993a940e39cb0a0b544ae8298546499b7e047
Summary:
A framework for tracing and replaying RocksDB operations.
A binary trace file is created by capturing the DB operations, and it can be replayed back at the same rate using db_bench.
- Column-families are supported
- Multi-threaded tracing is supported.
- TraceReader and TraceWriter are exposed to the user, so that tracing to various destinations can be enabled (say, to other messaging/logging services). By default, a FileTraceReader and FileTraceWriter are implemented to capture to a file and replay from it.
- This is not yet ideal to be enabled in production due to large performance overhead, but it can be safely tried out in a shadow setup, say, for analyzing RocksDB operations.
Currently supported DB operations:
- Writes:
-- Put
-- Merge
-- Delete
-- SingleDelete
-- DeleteRange
-- Write
- Reads:
-- Get (point lookups)
Pull Request resolved: https://github.com/facebook/rocksdb/pull/3837
Differential Revision: D7974837
Pulled By: sagar0
fbshipit-source-id: 8ec65aaf336504bc1f6ed0feae67f6ed5ef97a72
Summary:
The cache line size was computed dynamically based on the length of the filter bits, and the number of cache-lines encoded in the footer. This calculation had to be dynamic in case users migrate their data between platforms with different cache line sizes. The downside, though, was bloom filter probing became expensive as it did integer mod and division.
However, since we know all possible cache line sizes are powers of two, we should be able to use bit shift to find the cache line, and bitwise-and to find the bit within the cache line. To do this, we compute the log-base-two of cache line size in the constructor, and use that in bitwise operations to replace division/mod.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4071
Differential Revision: D8684067
Pulled By: ajkr
fbshipit-source-id: 50298872fba5acd01e8269cd7abcc51a095e0f61
Summary:
This adds support for writing unprepared batches based on size defined in `TransactionOptions::max_write_batch_size`. This is done by overriding methods that modify data (Put/Delete/SingleDelete/Merge) and checking first if write batch size has exceeded threshold. If so, the write batch is written to DB as an unprepared batch.
Support for Commit/Rollback for unprepared batch is added as well. This has been done by simply extending the WritePrepared Commit/Rollback logic to take care of all unprep_seq numbers either when updating prepare heap, or adding to commit map. For updating the commit map, this logic exists inside `WriteUnpreparedCommitEntryPreReleaseCallback`.
A test change was also made to have transactions unregister themselves when committing without prepare. This is because with write unprepared, there may be unprepared entries (which act similarly to prepared entries) already when a commit is done without prepare.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4104
Differential Revision: D8785717
Pulled By: lth
fbshipit-source-id: c02006e281ec1ce00f628e2a7beec0ee73096a91
Summary:
The member msgs of class Status contains all types of status messages.
When users dump a Status object, msgs will confuse users. So move it out
of class Status by making it as file-local static variable.
Closes#3831 .
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4144
Differential Revision: D8941419
Pulled By: sagar0
fbshipit-source-id: 56b0510258465ff26db15aa6b04e01532e053e3d
Summary:
Lint is not happy with some new code recently committed. Format them.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4161
Differential Revision: D8940582
Pulled By: siying
fbshipit-source-id: c9b43b1ef8c88b5e923911058b44eb77234b36b7
Summary:
Right now we use one hard-coded prefetch size to prefetch data from the tail of the SST files. However, this may introduce a waste for some use cases, while not efficient for others.
Introduce a way to adjust this prefetch size by tracking 32 recent times, and pick a value with which the wasted read is less than 10%
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4156
Differential Revision: D8916847
Pulled By: siying
fbshipit-source-id: 8413f9eb3987e0033ed0bd910f83fc2eeaaf5758
Summary: Windows requires new/delete for memory allocations to be overriden. Refactor to be less intrusive.
Differential Revision: D8878047
Pulled By: siying
fbshipit-source-id: 35f2b5fec2f88ea48c9be926539c6469060aab36
Summary:
Users sometime see their memtable size far smaller than expected. They probably have hit a fragementation of shard blocks. Cap their size anyway to reduce the impact of problem. 128KB is conservative so I don't imagine it can cause any performance problem.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4147
Differential Revision: D8886706
Pulled By: siying
fbshipit-source-id: 8528a2a4196aa4457274522e2565fd3ff28f621e
Summary:
Added Get Put Encode Decode support for Fixed16 (uint16_t). Unit test added in `coding_test.cc`
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4142
Differential Revision: D8873516
Pulled By: fgwu
fbshipit-source-id: 331913e0a9a8fe9c95606a08e856e953477d64d3
Summary:
Our "rocksdb.sst.read.micros" stat includes time spent waiting for rate limiter. It probably only affects people rate limiting compaction reads, which is fairly rare.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4102
Differential Revision: D8848506
Pulled By: miasantreble
fbshipit-source-id: 01258ac5ae56e4eee372978cfc9143a6869f8bfc
Summary:
The patch makes sure that two parallel test threads will operate on different db paths. This enables using open source tools such as gtest-parallel to run the tests of a file in parallel.
Example: ``` ~/gtest-parallel/gtest-parallel ./table_test```
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4135
Differential Revision: D8846653
Pulled By: maysamyabandeh
fbshipit-source-id: 799bad1abb260e3d346bcb680d2ae207a852ba84
Summary:
Picked up a task to convert this to use the gtest framework. It can't be this simple, can it?
It works, but should all the std::cout be removed?
```
[$] ~/git/rocksdb [gft !]: ./merge_test
[==========] Running 2 tests from 1 test case.
[----------] Global test environment set-up.
[----------] 2 tests from MergeTest
[ RUN ] MergeTest.MergeDbTest
Test read-modify-write counters...
a: 3
1
2
a: 3
b: 1225
3
Compaction started ...
Compaction ended
a: 3
b: 1225
Test merge-based counters...
a: 3
1
2
a: 3
b: 1225
3
Test merge in memtable...
a: 3
1
2
a: 3
b: 1225
3
Test Partial-Merge
Test merge-operator not set after reopen
[ OK ] MergeTest.MergeDbTest (93 ms)
[ RUN ] MergeTest.MergeDbTtlTest
Opening database with TTL
Test read-modify-write counters...
a: 3
1
2
a: 3
b: 1225
3
Compaction started ...
Compaction ended
a: 3
b: 1225
Test merge-based counters...
a: 3
1
2
a: 3
b: 1225
3
Test merge in memtable...
Opening database with TTL
a: 3
1
2
a: 3
b: 1225
3
Test Partial-Merge
Opening database with TTL
Opening database with TTL
Opening database with TTL
Opening database with TTL
Test merge-operator not set after reopen
[ OK ] MergeTest.MergeDbTtlTest (97 ms)
[----------] 2 tests from MergeTest (190 ms total)
[----------] Global test environment tear-down
[==========] 2 tests from 1 test case ran. (190 ms total)
[ PASSED ] 2 tests.
```
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4114
Differential Revision: D8822886
Pulled By: gfosco
fbshipit-source-id: c299d008e883c3bb911d2b357a2e9e4423f8e91a
Summary:
- Avoid `strdup` to use jemalloc on Windows
- Use `size_t` for consistency
- Add GCC 8 to Travis
- Add CMAKE_BUILD_TYPE=Release to Travis
Pull Request resolved: https://github.com/facebook/rocksdb/pull/3433
Differential Revision: D6837948
Pulled By: sagar0
fbshipit-source-id: b8543c3a4da9cd07ee9a33f9f4623188e233261f
Summary:
Copy data between buffers inside FilePrefetchBuffer only when chunk length is greater than 0. Otherwise AlignedBuffer was accessing memory out of its range causing crashes.
Removing the tracking of buffer length outside of `AlignedBuffer`, i.e. in `FilePrefetchBuffer` and `ReadaheadRandomAccessFile`, will follow in a separate PR, as it is not the root cause of the crash reported in #4051. (`FilePrefetchBuffer` itself has been this way from its inception, and `ReadaheadRandomAccessFile` was updated to add the buffer length at some point).
Comprehensive tests for `FilePrefetchBuffer` also to follow in a separate PR.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4100
Differential Revision: D8792590
Pulled By: sagar0
fbshipit-source-id: 3578f45761cf6884243e767f749db4016ccc93e1
Summary:
Right now slow deletion with ftruncate doesn't work well with checkpoints because it ruin hard linked files in checkpoints. To fix it, check the file has no other hard link before ftruncate it.
Pull Request resolved: https://github.com/facebook/rocksdb/pull/4093
Differential Revision: D8730360
Pulled By: siying
fbshipit-source-id: 756eea5bce8a87b9a2ea3a5bfa190b2cab6f75df
Summary:
Since the filter data is unaligned, even though we ensure all probes are within a span of `cache_line_size` bytes, those bytes can span two cache lines. In that case I doubt hardware prefetching does a great job considering we don't necessarily access those two cache lines in order. This guess seems correct since adding explicit prefetch instructions reduced filter lookup overhead by 19.4%.
Closes https://github.com/facebook/rocksdb/pull/4068
Differential Revision: D8674189
Pulled By: ajkr
fbshipit-source-id: 747427d9a17900151c17820488e3f7efe06b1871
Summary:
Currently, if RocksDB encounters errors during a write operation (user requested or BG operations), it sets DBImpl::bg_error_ and fails subsequent writes. This PR allows the DB to be resumed for certain classes of errors. It consists of 3 parts -
1. Introduce Status::Severity in rocksdb::Status to indicate whether a given error can be recovered from or not
2. Refactor the error handling code so that setting bg_error_ and deciding on severity is in one place
3. Provide an API for the user to clear the error and resume the DB instance
This whole change is broken up into multiple PRs. Initially, we only allow clearing the error for Status::NoSpace() errors during background flush/compaction. Subsequent PRs will expand this to include more errors and foreground operations such as Put(), and implement a polling mechanism for out-of-space errors.
Closes https://github.com/facebook/rocksdb/pull/3997
Differential Revision: D8653831
Pulled By: anand1976
fbshipit-source-id: 6dc835c76122443a7668497c0226b4f072bc6afd
Summary:
Various rearrangements of the cch maths failed or replacing = '\0' with
memset failed to convince the compiler it was nul terminated. So took
the perverse option of changing strncpy to strcpy.
Return null if memory couldn't be allocated.
util/status.cc: In static member function ‘static const char* rocksdb::Status::CopyState(const char*)’:
util/status.cc:28:15: error: ‘char* strncpy(char*, const char*, size_t)’ output truncated before terminating nul copying as many bytes from a string as its length [-Werror=stringop-truncation]
std::strncpy(result, state, cch - 1);
~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~
util/status.cc:19:18: note: length computed here
std::strlen(state) + 1; // +1 for the null terminator
~~~~~~~~~~~^~~~~~~
cc1plus: all warnings being treated as errors
make: *** [Makefile:645: shared-objects/util/status.o] Error 1
closes#2705
Closes https://github.com/facebook/rocksdb/pull/3870
Differential Revision: D8594114
Pulled By: anand1976
fbshipit-source-id: ab20f3a456a711e4d29144ebe630e4fe3c99ec25
Summary:
Previously in https://github.com/facebook/rocksdb/pull/3601 bloom filter will only be checked if `prefix_extractor` in the mutable_cf_options matches the one found in the SST file.
This PR relaxes the requirement by checking if all keys in the range [user_key, iterate_upper_bound) all share the same prefix after transforming using the BF in the SST file. If so, the bloom filter is considered compatible and will continue to be looked at.
Closes https://github.com/facebook/rocksdb/pull/3899
Differential Revision: D8157459
Pulled By: miasantreble
fbshipit-source-id: 18d17cba56a1005162f8d5db7a27aba277089c41
Summary:
Top-level index in partitioned index/filter blocks are small and could be pinned in memory. So far we use that by cache_index_and_filter_blocks to false. This however make it difficult to keep account of the total memory usage. This patch introduces pin_top_level_index_and_filter which in combination with cache_index_and_filter_blocks=true keeps the top-level index in cache and yet pinned them to avoid cache misses and also cache lookup overhead.
Closes https://github.com/facebook/rocksdb/pull/4037
Differential Revision: D8596218
Pulled By: maysamyabandeh
fbshipit-source-id: 3a5f7f9ca6b4b525b03ff6bd82354881ae974ad2
